Airborne Laser Altimetry Annotated Bibliography
LiDAR Bibliography:
Contents
Page
1. General/Introductory
2
2. Filtering and Strip Adjustment: Algorithms and Methods
4
3. DTM Generation; Terrain and Fluvial Applications
15
4. Calibration; Error Assessment; Quality Control
23
5. Commercial LiDAR
28
6. Forestry Applications
29
7. Urban Applications
34
8. Other Applications
40
9. Intensity
46
Last updated: 4/29/2004
Page 1 of 47
1. General/Introductory Title Airborne Laser Scanning – An Introduction and Overview
File Name
Author(s)
lidar_tutorial.pdf
Wehr, A., and U. Lohr
Publication and Annotated Abstract Journal of Photogrammetry and Remote Sensing 54: 68-82, 1999.
Abstract: The basic principles of LiDAR, the two main classes - pulse and continuous-wave lasers - and relations with respect to time-of-flight, range, resolution, and precision are presented. The main laser components and the role of the laser wavelength, including eye safety considerations, are explained. Different scanning mechanisms and the integration of laser with GPS and INS for position and orientation determination are presented. The data processing chain for producing digital terrain and surface models is outlined. Finally, a short overview of applications is given. Airborne Laser Scanning: Beyond its Formative Years
Ssc2003_Jonas.pdf
Jonas, D. and P. Byrne.
Available at: http://www.aamgeoscan.com.au/tech_list.htm AAM GeoScan, 2003.
Abstract: Features of ALS errors and the metrics which are used to express these errors, is discussed. Comparisons are drawn with other aerial surveying technologies, and concluding that conventional Gaussian errors statistics are inappropriate for ALS data. A convention for defining and expressing ALS errors is proposed which the author suggests is equally appropriate for other aerial surveying methods. Principles of Airborne Laser Scanning
ALSPrinciples_Kraus.pdf
Kraus, Karl
Journal of the Swedish Society for Photogrammetry and Remote Sensing, Nr. 2002:1, pp. 53 -56.
Abstract: This paper is the framework of a lecture presented in 2002 in Sweden on the occasion of the retirement of Prof. Dr. Torlegard. It is designed as an extended abstract with respective literature references from the Institute of Photogrammetry and Remote Sensing at the Technical University of Vienna (I.P.F.).
Airborne Laser Scanning
KK_ALS_2003.pdf
Kraus, Karl
Newsletter EuroSDR, 2003 No. 2, S. 11 - 12.
Abstract : In recent years, airborne laser scanning has become a dynamic branch of the applied sciences, research and development, and technology. The Institute of Photogrammetry and Remote Sensing at the Vienna University of Technology (I.P.F.) has been actively involved in this field for 8 years. Methods, algorithms and corresponding computer programs are developed for extracting information of laser scanner data, and pilot projects are performed for governmental institutions and private companies.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 2 of 47
Title Airborne Laser Scanning: Basic Relations and Formulas
File Name ALS_basicrelations.pdf
Author(s) Baltsavias, E.P.
Publication and Annotated Abstract ISPRS J. Photogrammetry and Remote Sensing, 54:199-214, 1999.
Abstract: The basic ALSM concepts and formulas are presented. The paper discusses general concepts of lasers and laser ranging and then focuses on pulse and CW airborne laser scanning. The factors influencing LiDAR data accuracy is discussed. Different scan patterns, especially parallel lines, are also discussed.
Airborne Laser Scanning: Supplying the Third Dimension
ALSintro_GeoScan.pdf
Aspects of Raster DEM Data Derived From Laser Measurements
Loeffler_ALSDD2003.pdf
Turton, David
AAM GeoScan, Australia. 2002. Available from: http://www.aamsurveys.com.au/tech_list.htm
Loffler, G.
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 810, 2003, Dresden, Germany
Abstract: Features of the TopoSys FALCON laser scanning system are described in the context of a general overview of raster DSM/DTM generation. Airborne Laser Scanning: a Broadacre Mapping Technique Airborne Laser Swath Mapping Shines New Light on Earth’s Topography
LiDAR Bibliography:
ALS_broadacreTechnique.pdf
AAM GeoScan
Slide Presentation. 48pp. AAM GeoScan, Australia. 2002. Available from: http://www.aamsurveys.com.au/tech_list.htm
Carter, W. Shrestha, R., Tuell, G., Bloomquist, D., Sartori, M.
EOS Transactions American Geophysical Union, 82:549, 550, 555, 2001.
Last updated: 4/29/2004
Page 3 of 47
2. Filtering and Strip Adjustment: Methods and Algorithms Title LiDAR Activities at the Viennese Institute of Photogrammetry and Remote Sensing
File Name rottensteiner_et_al.pdf
Author(s) Rottensteiner, F, Kager, H., Briese, C., and K. Kraus
Publication and Annotated Abstract Proceedings of the 3rd International LiDAR Workshop, "Mapping Geo-Surficial Processes Using Laser Altimetry", at the Ohio State University, Columbus, OH, October 7 - 9, 2002
Abstract: At the I.P.F, a method for filtering LiDAR data using robust linear prediction has been developed. In densely populated areas, the algorithm has to been used to eliminate large areas of points reflected from the roofs of buildings, which results in a hierarchic application of robust linear prediction. To improve the geomorphologic quality of DTMs, breaklines have to be detected. We describe two approaches for breakline detection. Hierarchic robust linear prediction can also be the basis for building extraction from raw LiDAR data: the points having been eliminated in DTM estimation are further classified according to whether they have been reflected off the tops of buildings, and connected regions of building points (“building regions”) are further analysed with the goal of finding planar patches and grouping these patches to reconstruct polyhedral building models. We also describe an approach to calibrate LiDAR data which involves simultaneous height fitting of LiDAR strips and a more sophisticated technique which models both the flight paths and the sensor attitudes by splines. Applications of the Robust Interpolation for DTM Determination
LiDAR Bibliography:
Briese_Pfeifer_Dorninger.pdf
Briese, Ch., Pfeifer, N., and P. Dorninger
ISPRS Commission III, Symposium. Sept., 2002, Graz, Austria
Abstract: We describe robust methods for the automatic generation of a DTM from point cloud data. These methods operate on the original data points and allow the elimination of non bare-earth points and the modelling of the terrain surface in a single step. We present results from the application of the algorithm to topographic data acquired by several different sensors (ALS, Terrestrial laser scanning, satellite laser scanning, tacheometry and photogrammetry).
Last updated: 4/29/2004
Page 4 of 47
Title Bayesian Object Recognition for the Analysis of Complex Forest Scenes in Airborne Laser Scanner Data
File Name
Author(s)
Publication and Annotated Abstract
BayesianRecognition.pdf
Andersen, H-E, Reutebuch, S., and G. Schreuder
ISPRS Commission III, Symposium. Sept., 2002, Graz, Austria. Available from: www.geo.unizh.ch/rsl/services/bibliographies/lidar/index.html
Abstract: Bayesian object recognition is applied to the analysis of complex forest geometry measured by high-density airborne laser scanning (LiDAR) data. With the emergence of high-resolution active remote sensing technologies, fine resolution, spatially distributed forest metrics can be extracted by statistical object recognition algorithms. A Bayesian approach to object recognition incorporates a probabilistic model of the active sensing process. LiDAR data is explicitly modelled in the domain of scan space, a three-dimensional analogue to two-dimensional image space. Prior models for object configurations take the form of Markov marked point processes, where pair-wise object interactions depend upon object attributes. The algorithm was applied to a 0.21 ha area within Capitol State Forest, WA, USA. Algorithm-based estimates are compared to photogrammetric crown measurements and field inventory data.
Derivation of Digital Terrain Models in the SCOP++ Environment
ALSfiltering_SCOP.pdf
Pfeifer, N. Stadler, P., and C. Briese
OEEPE Workshop on Airborne Laserscanning and Interferometric SAR for Digital Elevation Models, Stockholm, 2001
Abstract: We describe the development and application of a filtering
Filtering of Laser Altimetry Data
MorphologicalFiltering_Udelft.pdf
Vosselman, George
algorithm which has been used to generate bare-earth and canopy-oriented DTMs. The algorithm is an hierarchical, iterative robust linear prediction model. It adopts a coarse-to-fine approach which is both robust and computationally efficient. The results for OEEPE testing data are presented. Source: http://www.geo.tudelft.nl/frs/laserscan/filtering.html
Abstract: We describe an approach to filtering raw LiDAR data which uses training sets to determine the optimal filter function. We further assess the reliability of the generated DEM as a function of the point density.
Quality Improvement of Scanning Laser Altimeter Data
QualityImprovement.pdf
Behan, A.
Delft University of Technology, Section of Photogrammetry and Remote Sensing. June 2000. Available from: www.geo.tudelft.nl/frs/staff/avril/index.html
Abstract: Strategies for filtering ALS data using different surface generation routines are presented. A qualitative assessment identified significant differences between the DTMs generated by a TIN versus a nearest neighbor routine. The generated surface was less sensitive to grid cell size using the nearest neighbor algorithm. There were no observable differences for TIN-generated surfaces as a function of grid cell size. Based on this study, the author recommends TIN-based surface generation combined with block-adjustment filtering.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 5 of 47
Title Filtering Strategy: Working Towards Reliability
File Name Sithole03.pdf
Author(s) Sithole, George
Publication and Annotated Abstract ISPRS Commission III, Symposium 2002, September 9 - 13, Graz, Austria
Abstract: Measures to test LiDAR filter performance are typically generic indicators (RMSE, ratio of misclassifications vs. correct classifications, and so forth) which compare independently collected elevation data against filtered data. However, uncertainties remain with respect to local filter performance and which therefore call for alternative approaches which can distinguish different terrain and landtype features and adjust filter tolerances accordingly. Independently acquired data (satellite imagery, GIS data, and so forth) can be used both to guide bare-earth surface generation and assess the generated DTM quality. Poorly classified LiDAR data arises for chief three reasons: (1) the nature and arrangement of objects and surface terrain across the landscape (e.g., topography, buildings, vegetation, etc.,) (2) characteristics of the data (resolution, outliers, data gaps, etc.,) and, (3) the implementation of filters. We briefly discuss the first two and then present an approach to the third, which identifies areas where filters may perform poorly. Filtering of Laser Altimetry Data using a Slope Adaptive Filter
SlopeFilter_sithole.pdf
Sithole, George
In: Hofton, Michelle A. (ed) Proceedings of the ISPRS workshop on Land Surface Mapping en Characterization Using Laser Altimetry, 22nd to 24th October 2001, Annapolis, Maryland, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXIV part 3/W 4 Commission III. ISSN 0246 1840, pp.203-210
Abstract: We describe the development and application of a modified slope-based, morphological filter. The filter uses prescribed height thresholds to classify bare earth and non bare-earth surfaces. The filter was modified by incorporating slope-dependent (and thus, different terrain features) tolerances. This overcomes the limitation of the existing slopebased filter where tolerances are globally fixed, and thus limiting the effectiveness of the filter to areas of gentle relief. Results indicate that the modified filter reduces the number of Type I errors (non-bare earth objects in steep terrain that are not filtered). Few Type II errors are observed. We compare output of the modified slope filter, Terrascan, a commercial software filtering package.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 6 of 47
Title
File Name
Ground Surface Estimation from Airborne Laser Scanner Data Using Active Shape Models
ShapeModel_filtering.pdf
Airborne Laser Scanning: OEEPE Distance Learning Course
LinearPrediction_VUT.pdf
Author(s) Elmqvist, Magnus
Publication and Annotated Abstract ISPRS Commission III, Symposium 2002, September 9 - 13, Graz, Austria
Abstract: The development of an active shape model approach to estimate the ground surface from laser radar data is described. The active shape model acts like a rubber sheet which has both elasticity and rigidity. Using fixed parameters, the model generates an estimate of the bare earth surface. The algorithm is applicable to very dense data sets which can exceed 10 points per square meter. We discuss modifications to the algorithm which allows it to be applied to sparser data sets. Anon
Inst. of Photogrammetry and Remote Sensing, Vienna University of Technology. Downloaded from: http://www.ipf.tuwien.ac.at/OEEPE/tour/theory.htm
Abstract: The SCOP++ software package applies a linear prediction filtering algorithm to generate a bare-earth surface from raw laser data. The approach shares a similar theoretical foundation to kriging. Terrain heights are considered as a stochastic process - the realization of a spatial random variable. Observations (i.e. measurements of terrain height) and the stochastic properties (correlations) are used to estimate the value (i.e. the terrain height) of this random variable at different locations (i.e. ground plan positions, xy). Terrain Surface Reconstruction by the use of Tetrahedron Model with the MDL Criterion
TerrainReconstruct_dowman.pdf
Sohn, G., and I.J. Dowman
ISPRS Commission III, Symposium 2002, September 9 - 13, Graz, Austria
Abstract: We describe the development of a filtering algorithm which is used to extract a bare-earth surface from raw LiDAR data. A core feature of the approach is the implementation of an algorithm that continually adapts to local surface variability. Model development is described in terms of: (1) fragmentation of the topographic surface into piecewise “homogeneous” planar surfaces, where local surface roughness has been smoothed out, (2) a criterion for differentiating bare-earth and non-bare earth points, and (3) validation of our a priori assumptions about the final bare-earth surface. Several surfaces are generated iteratively and are used to test filter performance. The approach uses a two-step divide-and-conquer triangulation with downward and upward model refinement. A tetrahedron is used to approximate a terrain surface and the Minimum Description Length (MDL) criterion is used to select the optimized terrain surface model.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 7 of 47
Title Digital Terrain Models from Airborne Laser Scanning Data – A Grid-based Approach
File Name GridbasedFiltering_Wack.pdf
Author(s) Wack, R., and A. Wimmer
Publication and Annotated Abstract ISPRS Commission III, Symposium 2002, September 9 - 13, Graz, Austria
Abstract: We present an approach to LiDAR data filtering which combines a hierarchical strategy with a weighting function to detect non bare-earth elements. The weighting function considers the terrain shape as well as the distribution of the data points within a raster element. Algorithm of Multiple Filter to Extract DSM from LiDAR Data
Algorithm of Multiple Filter to Extract DSM from LiDAR data.pdf
Okagawa, M.
ESRI Users Conference, 2001
Abstract: LiDAR data is classified using the neighborhood height effect and a clustering criterion. Surface Clustering from Airborne Laser Scanning Data
SurfaceClustering.pdf
Filin, S.
ISPRS Commission III, Symposium. Sept., 2002, Graz, Austria.
Abstract: An algorithm for the extraction of surface clusters from airborne laser data is presented. Surface structure analysis is fundamental to almost any application involving LiDAR data, yet most algorithms focus only on identifying planar segments. The proposed algorithm is more general insofar as it aims to extract surface segments that exhibit homogeneous behavior without restriction to a specific pattern. The algorithm adopts a data clustering methodology which provides a general and flexible way to identify homogeneous patterns in the data. Object Segmentation with Region Growing and Principal Components Analysis
Laser Strip Adjustment for Data Calibration and Verification
LiDAR Bibliography:
ObjectsSegmentation_PCA.pdf
Roggero, M.
ISPRS Commission III, Symposium. Sept., 2002, Graz, Austria.
Abstract: The paper considers the problem of object segmentation and shape recognition in discrete, noisy data. Two algorithms, one which uses region growing strategy, and the other, principal component analysis, are used to filter ALS data. StripAdjustment.pdf
Burman, H.
ISPRS Commission III, Symposium. Sept., 2002, Graz, Austria.
Abstract: Laser scanning relies on satellite positioning and inertial navigation for accurate georeferencing. Georeferencing errors are a result of: (1) satellite positioning error (e.g., atmospheric delay), (2) cycle slips and loss-of-lock together with drifts in accelerometers, and, (3) gyros in the inertial system resulting in systematic orientation errors. Some errors can be corrected by matching and groundtruthing overlapping laser strips. This paper describes development and testing of a laser strip adjustment program, TerraMatch.
Last updated: 4/29/2004
Page 8 of 47
Title
File Name
On the Estimation of Planimetric Offsets in Laser Altimetry Data
PlanimetricOffsets.pdf
Adjustment and Filtering of Raw Laser Altimetry Data
OEEPE_Stockholm.pdf
Author(s) Vosselman, G.
Publication and Annotated Abstract ISPRS Commission III, Symposium. Sept., 2002, Graz, Austria.
Abstract: Offsets between overlapping ALS strips are used to estimate and eliminate systematic errors in laser altimetry data. For 3D strip adjustment, offsets are measured in three dimensions. Height offsets can be determined by comparing the heights of horizontal planes. Planimetric offsets are more difficult to determine. We show that using least squares algorithms on height data as well as on reflectance data, can lead to significant biases in the estimation of planimetric offsets. For height data, we propose a model based on estimation of linear features. To improve both the offset estimation and variance estimation with the reflectance data, an edge response function is introduced. The function accounts for the difference in size of a laser beam's footprint and the distance between successive laser points. Vosselman, G., and H-G Mass
Workshop on laser scanning and interferometric SAR for digital elevation models. Stockholm, 2001.
Abstract: Two important limitations of LiDAR data are: (1) the detection of, and correction of, systematic strip errors, and (2) filtering bare-earth data from non bare-earth objects. Systematic errors can be broadly categorized as: (a) strip wise errors, which are a function of GPS/INS imperfections, and (b) local errors, which are a function of the laser system and GPS noise. We use a least squares matching (LSM) model within a TIN to remove both these error sources. Filtering is performed using a mathematical morphological algorithm which uses slope-based thresholds to classify bareearth and non bare-earth objects. In urban areas, independently acquired GIS data significantly improves surface generation. The authors argue that LiDAR data processing should proceed first with strip adjustment followed by filtering. Processing of Laser Scanner Data – Algorithms and Applications.
LiDAR Bibliography:
lidar_algorithms.pdf
Axelsson, P.
J. Photogrammetry and Remote Sensing, 54: 138-147, 1999
Abstract: Filtering algorithms for determining the bare-earth surface, classification of buildings for 3D City Models, and the detection of electrical power lines, are presented. The algorithms use the Minimum Description Length (MDL) criterion. Additional use of reflectance data and multiple echoes are found to be useful for several applications.
Last updated: 4/29/2004
Page 9 of 47
Title Some Algorithms for Virtual Deforestation (VDF) of LiDAR Topographic Survey Data
File Name haugerud_algorithm.pdf
Author(s) Haugerud, R. A., and D.J. Harding
Publication and Annotated Abstract Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: We describe a virtual deforestation (VDF) despike algorithm that classifies returns as ground or not-ground on the basis of the geometry of the surface in the neighborhood of each return. The algorithm is fully automatic, effective, and can recover breaklines. It fails to identify some negative blunders, rounds some sharp corners off the landscape, and as implemented is slow. If multiple-return data are available, a no-multiplereturns VDF algorithm robustly defines areas where all returns are ground returns. Many groups are using variations on block-minimum VDF algorithms, but these do not work well on slopes and typically require substantial manual effort to adjust block size as the fraction of ground returns changes. Fully automatic VDF algorithms are desirable not only to minimize survey costs but also to produce topography for which all necessary interpretive biases and assumptions are explicit. The development of effective VDF algorithms has been hindered by the tendency of some commercial and academic practitioners to keep their work proprietary. Open dialogue is needed. A Top-Down Operator for the Automatic Extraction of Trees: Concept and Performance Evaluation
Straub_ALSDD2003.pdf
Straub, B-M.
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: A multi-scale approach for automatic LiDAR data extraction is presented. The segmentation of the surface model is performed using a watershed transformation at multiple scales for the purpose of detecting trees in different size classes. Four parameters are used for tree detection. Three are object-specific, while the fourth depends on the raster data. We show results for urban and forest environments. Managing and Processing LiDAR Data within GRASS
Brovelli_Maria_Antonia.pdf
Brovelli, M.A., M. Cannata, and U.M. Longoni
Proceedings of the Open Source GIS-GRASS users conference, 2002. Trento, Italy.
Abstract: Automatic extraction of bare earth DTMs from raw LiDAR data is implemented in the grid-based GIS package, GRASS. An initial surface is interpolated using a bicubic spline. Histogram analysis of the residuals are used to evaluate outlier elimination thresholds.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 10 of 47
Title Classification and Filtering of Laser Data
File Name
Author(s)
Nardinocchi_ALSDD2003.pdf
Nardinocchi, C., Forlani, G., and P. Zingaretti
Publication and Annotated Abstract Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: We present a strategy for raw LiDAR filtering which uses a 3tier classification based on geometric and topological criteria. The three classes are: urban, vegetation, and terrain. Following rasterization of the raw data, a region-growing algorithm which uses slope thresholds, is applied to fill in data gaps. Raw data are rasterized and segmented in connected regions that are bordered by a step edge. Data gaps, vegetation regions, and data errors are identified using size and region fragmentation criteria. Grid interpolation and raw data filtering are work in progress. Filtering of Laser Scanning Data in Forest Areas Using Finite Elements
Krzystek_ALSDD2003.pdf
Krzystek, P.
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: LiDAR data is processed by combining a pre-filter based on a convex hull and subsequent finite element adjustment. The pre-filter creates a crude TIN which assumes that the lowest laser points represent the bareearth surface. The adjustment of the TIN uses all laser points except nonground points. This leads to a slight height change of the TIN points according to statistical and geometrical constraints and to a better approximation of the terrain. The method works hierarchically in a data pyramid starting from a coarse point density of about 4 m2. The approach has been applied to data from the National Park Bavarian Forest using the TopoSys laser system. Results indicate that the method is appropriate to forested areas with different vertical forest structures. Vertical accuracy ranges from 15 – 25 cm. Determination of Terrain Features in a Terrain Model from Laser Radar Data
Lantz_ALSDD2003.pdf
Lantz, F., Jungert, E., and M. Sjovall
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: A terrain model that determines terrain features using symbolic filters is presented. The filters are compositions of symbolic surface elements.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 11 of 47
Title
File Name
A Hybrid Model for DTM Generation from LiDAR Signatures
lee_phd.pdf
Author(s) Lee, Hyun.S.
Publication and Annotated Abstract Unpublished PhD dissertation submitted to the Department of Electrical and Computer Engineering, Mississippi State University. December, 2003.
Abstract: An approach to extract bare earth elevation from LiDAR data is presented. The technique involves a preprocessing step, ground modeling, and interpolation. In the preprocessing step, most of the non-terrain points are eliminated using a histogram-based clustering technique. In the ground modeling stage, information such as elevation and slope between nearest neighbor points is extracted. This step represents an outlier detection process. Residuals and gradient indices for elevation and slope, are introduced. These indices discard remaining non-terrain points. The final DTM is generated using splines. Results show an improvements over existing techniques. Prospects of using a ground trend model developed from specific tree height measurements is also discussed. A Progressive Morphological Filter for Removing No ground Measurements from Airborne LiDAR Data
TGRS.pdf
Derivation of Range and Range Distributions from Laser Pulse Waveform Analysis for Surface Elevations, Roughness, Slope, and Vegetation Heights
waveform_9_03.pdf
LiDAR Bibliography:
Zhang, K., and 5 other authors
IEEE Trans. Geoscience and Remote Sensing, Vol. 41(4), April 2003
Abstract: A progressive morphological filter to detect non-ground LiDAR points is presented. By increasing the window size of the filter and using elevation difference thresholds, non bare-earth data, including vehicles, vegetation, and buildings, are removed, while preserving bare earth data. Datasets from mountainous and flat urbanized areas were selected to test the filter. The approach involves gridding and then the raw LiDAR. Some data loss accompanies the gridding process which can be limited by gridding at finer resolutions. However, the computational costs increase significantly with smaller grid cells, while the constraints imposed by using a regularlyspaced grid remain. Negative blunders are also not handled by the filter. The authors report filter performs satisfactorily for two test sites. The authors note that parameterization is crucial for satisfactory filter performance, and also acknowledge that integrating other sources of information will likely improve filter performance. Brenner, A.C. and 10 other authors
Geoscience Laser Altimeter System (GLAS), Nasa Goddard Space Flight Center, U. Wisconsin. September, 2003.
Abstract: GLAS detects ice elevation changes by precision profiling of ice surface elevations over the Greenland and Antarctic ice sheets. It also measures sea ice, ocean and land surface elevations; ice, water, and land surface roughness; multiple near-surface canopy heights over land; and cloud and aerosol layer heights.
Last updated: 4/29/2004
Page 12 of 47
Title Segmented Filtering of Laser Scanner DSMS
File Name Jacobsen_ALSDD2003.pdf
Author(s) Jacobsen, K., and P. Lohmann
Publication and Annotated Abstract Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: Limitations of automatic global filtering DSM strategies are described. Raw LiDAR data is classified according to land use. Filter tolerances and thresholds are evaluated automatically for each land-use, resulting in optimal, spatially variable filtering. Filtering of Digital Elevation Models
jac_Filtasp.pdf
Jacobsen, K, and R. Passini
Proceedings of FIG / ASCM / ASPRS annual convention, Washington. 2002
Abstract: We describe and present results from application of a linear prediction algorithm combined with a stationary random function after trend removal which was used to filter raw LiDAR data. Filter performance was assessed for different surface types, including, urban areas, forested terrain, variable roughness, and for different grid cell resolution. Surface Estimation Based on LiDAR
Article_SurfaceEstimation.pdf
Schickler, W, and A. Thorpe
Proceedings of the ASPRS Annual Conference, St. Louis, MO, April 2001
Abstract: We describe an approach for DTM extraction which uses raw LiDAR data and ancillary information, including independently acquired breaklines and surface categories. We apply a least-squares adjustment with robust estimation similar to that proposed by (Kraus, Pfeifer, 1998).
LiDAR Bibliography:
Last updated: 4/29/2004
Page 13 of 47
Title ISPRS: Comparison of Filters
File Name
Author(s)
Report05082003.pdf
Sithole, G., and G. Vosselmann
Publication and Annotated Abstract ISPRS Commission III, Working Group 3 Report, August 2003
Abstract: Raw LiDAR data processing involves modeling of systematic errors, filtering, feature detection, and thinning. Filtering and quality control tests can consume 60-80% of processing time. We examine the performance of several filters using a synthetic terrain surface. Seven features of the filters were identified: (1) data structure, (2) test neighbourhood, (3) measure of discontinuity, (4) filter concept, (5) single vs. iterative processing, (6) replacement vs. culling, and (7) use of first pulse and reflectance data. All the filters were challenged by areas of complex terrain, including buildings on slopes, disconnected terrain (courtyards), and the preservation of discontinuities. Fifteen sub samples were extracted from the eight data sets. A quantitative comparison of the various filter performance was based on results for 15 sample areas. The filters generally performed satisfactorily for ‘simple’ terrain. However, generating a bare-earth surface in complex terrain posed significant challenges. Future research should be directed towards heuristic classification of point-clouds (based on external data), quality reporting, and improving the filter efficiency. Slope Based Filtering of Laser Altimetry Data
vosselmanfiltering.pdf
Vosselman, G.
IAPRS, Vol. XXXIII. Amsterdam, 2000.
Abstract: We present an approach to filtering LiDAR data that applies a method related to the erosion operator used in mathematical gray scale morphology. Based on height differences in training datasets, filter functions are evaluated that either preserve important terrain characteristics or minimise the number of classification errors. The ‘minimisation’ filter results in fewer errors in the generated DTM.. The performance of the filter deteriorates with a decreasing point density. A Method To Predict Accuracy of Least Squares Matching For Airborne Laser Scanning Data Sets
Paquet_ALSDD2003.pdf
Paquet, R.
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: A surface matching algorithm which incorporates different source data into the same reference system, is presented. Vertical and horizontal georeferencing accuracies are shown to be strongly dependent on raw data density. We show that the curves generated by a cross-validation method can be modelled using hyperbolic functions. The accuracy of the predictions ranged from sub-millimetre values to 4mm for both tests. Registration accuracy of the ALS data where the differences ranged from 9mm to 134mm, was the exception. In this case however the accuracy improved with increasing density.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 14 of 47
3. DTM Generation; Terrain and Fluvial Applications Title Surface Estimation Based on LiDAR
File Name Article_SurfaceEstimation.pdf
Author(s) Schickler, W, and A. Thorpe
Publication and Annotated Abstract Proceedings of the ASPRS Annual Conference, St. Louis, MO, April 2001
Abstract: We present an approach for bare-earth estimation which uses independent data to guide the filtering process. The supplemental data includes independently measured breaklines and surface landtypes information. The filtering algorithm is a least-squares adjustment with robust estimation similar to that proposed by (Kraus, Pfeifer, 1998). The surface model is represented using a triangular irregular network or TIN. Airborne Laser Scanning and Derivation of Digital Terrain Models
ALS_to_DTM.pdf
Briese, C., and N. Pfeifer
In Grün/Kahmen (Eds.): Optical 3-D Measurement Techniques V, 2001, pp. 80-87.
Abstract: Development and testing of SCOP++, an iterative robust interpolation algorithm (using linear prediction) which is being used to filter bare-earth information in forested and urban environments is presented. Results are satisfactory but could be improved by using independent GIS data, which would include breaklines and other important features. The potential for using first and last pulse returns along with their intensities, to better characterize the bare-earth surface, is discussed. Semantically Correct Integration of a Digital Terrain Model and a 2D Topographic Vector Data Set
Evaluation of Light Detection and Ranging (LIDAR) for Measuring River Corridor Topography
LiDAR Bibliography:
Kock0309stutt.pdf
Koch, A.
Proceedings of ISPRS Workshop on Challenges in Geospatial Analysis, Integration and Visualization II, Stuttgart, 8.-10. Sept. 2003, CD
Abstract: We describe an algorithm which uses semantic information about an object’s properties to more accurately integrate raster DTMs and 2D vector topographic data. The algorithm is based on an inequality constrained least squares adjustment formulated as the linear complementary problem (LCP). Results are presented for a simulated dataset representing a tilted plane and for real, surveyed data. Bowen, Z.H, R.G. Waltermire
Journal of the American Water Resources Association, 38(1): 33-41. 2002
Last updated: 4/29/2004
Page 15 of 47
Title The Effect of Changing Grid Size in the Creation of Laser Scanner Digital Surface Models
File Name Smith_Paper.pdf
Author(s) Smith, S.L., Holland, D.A., and Longley, P.A.
Publication and Annotated Abstract Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 810, 2003, Dresden, Germany
Abstract: Limited understanding of the magnitude and the spatial distribution of errors results leads to uncertainties regarding the generated DTM and DSMs, and also uncertainty in processes such as object reconstruction and viewshed modelling. For object reconstruction and feature extraction in particular, understanding the magnitude, pattern and characteristics of the errors introduced by different interpolation methods is vital. This paper presents the results of a sensitivity analysis of the effects of varying the resolution of resampled locations upon the magnitude of errors. Three grid sizes are investigated, and the magnitudes and spatial pattern of errors within each are identified. 45 DSMs were generated using five interpolation algorithms and 3 different grid cell sizes. Laser altimeter return pulse correlation: a method for detecting surface topographic change
LiDAR Bibliography:
pulsecorrelation.pdf
Hofton, M.A., and J.B. Blair
Journal of Geodynamics 34 (2002) 477–489 Abstract: Quantifying and monitoring of many natural hazards requires repeated measurements of a topographic surface whose change reflects a geological or geophysical process. We evaluate the feasibility of a method for using laser altimeter return echoes, or waveforms, to detect relative elevation change. The method, the return pulse correlation method, maximizes the shape similarity of coincident laser return waveforms from two observation epochs by shifting them vertically. We evaluate the accuracy of the pulse correlation method using LiDAR data acquired over the NASA Wallops Flight Facility, VA, a region where no elevation change is expected within the time period of the surveys, and at Assateague Island, MD, a highly dynamic barrier island where several meters of erosion and deposition have been observed. Results show that the method generates elevation change estimates similar to those obtained by simply differencing coincident laser altimeter elevation measurements (the spot comparison method). Along the beach at Assateague Island, MD, similar patterns of accretion and deposition are detected using both the pulse correlation and spot comparison methods, although some horizontal resolution is lost using the pulse correlation method because of the wide footprint spacing of the waveform-recording laser altimeter. Increasing the size of the laser footprint from 25 to 60 m caused the magnitude of the vertical change signal to be underestimated, indicating that the resolution of the measurement technique and the scale of the deformation features should be considered when planning survey missions.
Last updated: 4/29/2004
Page 16 of 47
Title Automated analysis and classification of landforms using high-resolution digital elevation data: applications and issues
File Name Automated_geohydro.pdf
Fusion of LiDAR Data and Aerial Imagery for a More Complete Surface Description
AerialPhotoFusion_schenk.pdf
Determination of Terrain Models in Wooded Areas with Airborne Laser Scanner Data
alsm_wooded.pdf
LiDAR Bibliography:
Author(s) MacMillan, T.C. Martin, T.J. Earle, and D.H. McNabb
Schenk, T., and B. Csatho
Publication and Annotated Abstract Can. J. Remote Sensing, Vol. 29, No. 5, pp. 592–606, 2003
Abstract: This paper describes recent efforts to use fine spatial resolution digital elevation model (DEM) data, including, but not limited to, lidar DEM data, for automated analysis and classification of geomorphic and hydrologic terrain features. The example applications presented are based on a 3-m horizontal resolution lidar DEM for a 6 km by 8 km agricultural watershed in Alberta and on a similar 5-m horizontal resolution conventional DEM for two forested areas of 14 km by 12 km in the Cariboo Forest Region of British Columbia. The applications illustrate efforts to produce meaningful classifications of ecological and landform spatial entities and to automatically extract hydrological spatial entities required for input into the WEPP water erosion model. Fine spatial resolution DEMs present some unique problems for which solutions are still lacking or are insufficient. Errors of apparently minor extent or degree can seriously affect some forms of analysis, especially analyses that involve hydrological calculations of paths of surface water flow. A need is recognized for improved methods and tools for interpolating and editing fine spatial resolution DEMs to remove or reduce localized errors. ISPRS Commission III, Symposium 2002, September 9 - 13, Graz, Austria
Abstract: This paper describes two aspects of merging aerial imagery and LiDAR data. Establishing a common reference frame is crucial and is handled by utilizing sensor-invariant features (e.g., breaklines). We discuss synergisms between these features results in a richer and more abstract surface description. Kraus, K., and N. Pfeifer
ISPRS J. Photogrammetry and Remote Sensing, 53: 193-203. 1998.
Abstract: We compare DTMs generated by ALS versus photogrammetrically generated DTMs. ALS data are characterized by a skewed error distribution function due to non-bare earth objects being included in the original raw dataset. We describe an adaptive filtering strategy which uses linear prediction and automatically extracts breaklines. The filter performs satisfactorily. However, contours interpolated from the DTM are at best, reasonable. The final DTM will be improved by using independent information (e.g., breaklines) to help guide filtering.
Last updated: 4/29/2004
Page 17 of 47
Title Advanced DTM Generation from LiDAR Data
File Name AdvancedDTMgen.pdf
Author(s) Kraus, K., and N. Pfeifer
Publication and Annotated Abstract International Archives of Photogrammetry and Remote Sensing, Volume XXXIV-3/W4, Annapolis, Maryland, 22. - 24. October, 2001, pp 23-30.
Abstract: A technique for calibrating laser scanner data is introduced. Height discrepancies between overlapping strips and ground control points are used to transform all strips into a statewide coordinate system. A bare-earth surface is then generated from the raw ALS data. Filtering and terrain interpolation are performed simultaneously which results in more accurate classification of bare earth data in steep terrain. Generating accurate DTMs also requires that breaklines be used to guide filtering. The approach incorporates surface hydrological concepts to guide DTM generation. Flow paths and sinks are identified in a preliminary DTM. The revised DTM incorporates this information by defining drainage lines and sinkfilling. We also extract 3D breaklines if the approximate location of the breaklines in the original ALS data is known. The result of this step are 3D-splines which are integrated in the hybrid DTM, combining raster and vector data. Automated Analysis and Classification of Landforms using HighResolution Digital Elevation Data: Applications and Issues
The Effect of LiDAR Posting Density on DEM Accuracy and Flood Extent Delineation: A GIS-Simulation Approach
LiDAR Bibliography:
automated_geohydro.pdf
Georgeraber_phd.pdf
MacMillan, R.A., Martin, T.C., Earl, T.J., and D.H. McNabb
Raber, George
Canadian Journal of Remote Sensing, Vol. 29(5): 592-606, 2003.
Abstract: Recent efforts to apply fine spatial resolution DEMs for automated analysis and classification of geomorphic and hydrologic terrain features are described. The examples presented use a 3-m horizontal resolution lidar DEM for a 6 km by 8 km agricultural watershed in Alberta and a 5m ‘conventional’ DEM for two forested areas of 14 km by 12 km in the Cariboo Forest Region of British Columbia. The applications illustrate efforts to produce meaningful classifications of ecological and landform spatial entities and to automatically extract hydrological spatial entities required for input into the WEPP water erosion model. Fine spatial resolution DEMs present some unique problems for which solutions are still lacking or are insufficient. Errors of apparently minor extent or degree can seriously affect some forms of analysis, especially analyses that involve hydrological calculations of paths of surface water flow. A need is recognized for improved methods and tools for interpolating and editing fine spatial resolution DEMs to remove or reduce localized errors. Unpublished paper submitted as part of graduate research, NASA Affiliated Research Center, Dept of Geography, U. South Carolina.
Abstract: An empirical relationships between LiDAR posting density and DEM accuracy is presented. The relationship can be used in future LiDAR projects to guide data collection and processing efforts in order to meet predefined accuracy. ALS data sampling for different posting densities are simulated for 2 synthetic surfaces. Vertical accuracy and angular error statistics from the generated surfaces are compared to original data posting density. Both metrics are strongly linearly dependent on posting density.
Last updated: 4/29/2004
Page 18 of 47
Title 3D Terrain Models on the Basis of a Triangulation
File Name
Author(s)
PfeiferN_PhD.pdf
Pfeifer, Norbert
Publication and Annotated Abstract PhD dissertation, Institut für Photogrammetrie und Fernerkundung, Technischen Universität Wien. 2002. Abstract: Several approaches for surface generation using triangulation are discussed. Two approaches are highlighted: (1) the patch work, and (2) the subdivision approach. Surfaces generated by the subdivision approach better meet the requirements for DTM generation.
Objective Landslide Detection and Surface Morphology Mapping using High-Resolution Airborne Laser Altimetry
LiDAR Bibliography:
2003Geomorphology-McKean.pdf
Mckean, J., and J. Roering
Geomorphology, Volume. 1412: 1-21. 2003.
Abstract: We constructed accurate, high-resolution DEMs from airborne laser altimetry (LiDAR) data to characterize a large landslide complex and surrounding terrain near Christchurch, New Zealand. One-dimensional, circular (2-D) and spherical (3-D) statistics are used to map the local topographic roughness in the DEMs over a spatial scale of 1.5 to 10 m. The bedrock landslide is rougher than adjacent unfailed terrain and any of the statistics can be employed to automatically detect and map the overall slide complex. Furthermore, statistics that include a measure of the local variability of aspect successfully delineate four kinematic units within the gently sloping lower half of the slide. Features with a minimum size of surface folds that have a wavelength of about 11 to 12 m and amplitude of about 1 m are readily mapped. Two adjacent earthflows within the landslide complex are distinguished by a contrast in median roughness, and texture and continuity of roughness elements. The less active of the earthflows has a surface morphology that presumably has been smoothed by surface processes. The Laplacian operator also accurately maps the kinematic units and the folds and longitudinal levees within and at the margins of the units. Finally, two-dimensional power spectra analyses are used to quantify how roughness varies with length scale. These results indicate that no dominant length scale of roughness exists for smooth, unfailed terrain. In contrast, zones with different styles of landslide deformation exhibit distinctive spectral peaks that correspond to the scale of deformation features, such as the compression folds. The topographic-based analyses described here may be used to objectively delineate landslide features, generate mechanical inferences about landslide behavior, and evaluate relatively the recent activity of slides.
Last updated: 4/29/2004
Page 19 of 47
Title
File Name
A Network-Index-Based Version of TOPMODEL for use with HighResolution Digital Topographic Data
TOPMODEL_lidar04.pdf
Terrain Based Descriptions of Earth Surface and Atmospheric Processes
source_dtm_comparison.pdf
Author(s) Lane, S.N., Brookes C.J., Kirkby M.J., and J. Holden
Wilson, J.P.
Publication and Annotated Abstract Hydrologic Processes, 18: 191-201, 2004
Abstract: This paper describes the preliminary development of a network-index approach to modify and to extend the classic TOPMODEL. Application of the basic Beven and Kirkby form of TOPMODEL to high-resolution (2·0 m) laser altimetric data (based upon the UK Environment Agency’s light detection and ranging (LIDAR) system) to a 13·8 km2 catchment in an upland environment identified many saturated areas that remained unconnected from the drainage network even during an extreme flood event. This is shown to be a particular problem with using high-resolution topographic data, especially over large appreciable areas. To deal with the hydrological consequences of disconnected areas, we present a simple network index modification in which saturated areas are only considered to contribute when the topographic index indicates continuous saturation through the length of a flow path to the point where the path becomes a stream. This is combined with an enhanced method for dealing with the problem of pits and hollows, which is shown to become more acute with higher resolution topographic data. The paper concludes by noting the implications of the research as presented for both methodological and substantive research that is currently under way. GeoSpatial World 2002, Conference Proceedings. Atlanta, Georgia. June 10-12, 2003. Available from: http://www.geospatialworld.com/gsw2002/proceedings2002/default.asp
Abstract: Terrain analysis applications are discussed. Algorithms used in current terrain analysis software, algorithm performance, interpretation of terrain attributes, use of terrain attributes in predictive models, and the effects of scale and resolution are also discussed. Objective Landslide Mapping and Surface Morphology Mapping using High-Resolution Airborne Laser Altimetry
RoeringMcKean_geomorph04.pdf
McKean, J. and J. Roering
Geomorphology 57: 331-351, 2004.
Abstract: The authors present an alternative approach to landslide mapping and inventory for a deep-seated landslide complex near Christchurch, New Zealand. 1D, circular 2D, and spherical 3D geostatistics are used to map topographic roughness over spatial scales ranging from 1.5 to 10 meters. The statistical techniques affectively distinguish between slide and non-slide areas and furthermore, measures of aspect distinguish 4 different kinematic units. Less active kinematics are characterized by smoother surfaces. Application of high resolution topographic
data serves to more effectively delineate the spatial extent of landslide features while providing indirect information on process mechanics and the timing of landslide activity.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 20 of 47
Title Airborne LiDAR in Support of Geomorphological and Hydraulic Modelling
Application of Airborne Lidar in River Environments: the River Coquet, Northumberland, UK.
LiDAR Bibliography:
File Name french_2003espl.pdf
Author(s) French, J.N.
Publication and Annotated Abstract Earth Surface Processes and Landforms, 28: 321-335. 2003
Abstract: This paper considers the application of airborne laser altimetry (LiDAR) to the provision of elevation data at accuracies and spatial densities commensurate with the current generation of high-resolution hydraulic models. Three sets of issues are addressed with reference to a Telemac 2D model of a morphologically complex estuary in eastern England. First, the quality of airborne LiDAR data is assessed via multiscale calibration against surveyed sections and supplementary control points. Second, image processing techniques are used for (i) identification of multiple regions of interest within large LiDAR mosaics; (ii) subregional infilling of voids left by data ‘drop-outs’; (iii) filtering and subsampling to match topographic information content with model resolution and the density of the computational mesh. Third, the implications of improved terrain data are considered with reference to the estimation of elevations and potential tidal volumes for a number of discrete flood compartments and the modelling of hypothetical inundation scenarios. After minor offset correction to ensure registration with local benchmarks, the quality of LiDAR elevation data within a 12 km ð 4 km coverage is found to be consistent with the published specification. The accuracy and spatial resolution of the LiDAR data allow the identification of subtle but important topographic variations between adjacent flood compartments. In many cases, differences in model results obtained using these data relative to previously estimated average flood compartment elevations are small. Importantly, however, LiDAR provides topographic information at an accuracy and resolution close to the present limits of model representation. Reliable representation of form allows the modeller to concentrate on the physical aspects of model parameterization, whilst minimizing the conflation of parameter effects with those of poorly constrained geometric representation. charlton2003_espl.pdf
Charlton, M.E, A.R.G. Large, and I.C. Fuller
Earth Surface Processes and Landforms, 28: 299-306. 2003
Abstract: The potential offered by LiDAR (laser-induced direction and ranging) for the mapping of gravel-bed river environments is addressed in this paper. A LiDAR dataset was obtained for a reach of the River Coquet, Northumberland, UK. Topographic data were acquired from the field at the same time using theodoliteEDM survey of a number of cross-profiles across the active river channel and bar units. These cross-profiles provide a means of comparing measurements from the LiDAR data with ground survey. Ordnance Survey large-scale mapping was used to georeference the survey data, which were then integrated with the LiDAR dataset using GIS software. A close correspondence between ground survey-derived crossprofiles and those generated using LiDAR is observed. However, the presence of both vegetation and deep water introduces anomalies in the LiDAR surface. Correction for these anomalies is needed to improve the accuracy of LiDAR mapping in the UK context and similar river environments. It is concluded that LiDAR has potential as an accurate survey tool for obtaining high-resolution topographic data from unvegetated, exposed bar surfaces.
Last updated: 4/29/2004
Page 21 of 47
Title LiDAR Data Filtering and DTM Interpolation Within Grass
LiDAR Bibliography:
File Name LidarFiltering_GIStrans2004.pdf
Author(s) Brovelli, M.A., Cannata, M., and Ulisse M. Longoni
Publication and Annotated Abstract Transactions in GIS, 2004, 8(2): 155-174 LIDAR is based on the combination of three different data collection tools: a laser scanner mounted on an aircraft, a Global Positioning System (GPS) used in phase differential kinematic modality to provide the sensor position and an Inertial Navigation System (INS) to provide the orientation. For standard conditions, taking into account flight speed (200–250 km/h, altitude 500–2,000 m) and sensor characteristics (scan angle ± 10–20 degrees, emission rate 2–50 KHz), elevation data are acquired with a density of one point per 0.5–3 m. We present a procedure for bare-earth data extraction which tasks advantage of intensity information. The approach extracts an initial bare-earth surface using a bicubic spline interpolation and least-squares adjustment, which serve to remove data outliers.
Last updated: 4/29/2004
Page 22 of 47
4. Calibration, Error Assessment, Quality Control Title
File Name
Huising, E.J., Gomes Pereira, L.M.
Errors and accuracy estimates of laser data acquired by various laser scanning systems for topographic applications
Quality Assessment of Laser-Scanned Data
Author(s)
isprs99%20koch20lohmann.pdf
Lohmann, P. and A. Koch
Publication and Annotated Abstract ISPRS J. Photogrammetry and Remote Sensing, 53:245-261, 1998.
Abstract:: The Survey Department of Rijkswaterstaat in The Netherlands makes extensive use of laser scanning for topographic measurements. An inventory of sources of errors indicates that errors may vary from 5 to 200 cm. The experience shows that errors related to the laser instrument, GPS and INS may frequently occur, resulting in local distortions, and planimetric and height shifts. Moreover, the results indicate that for flat terrain, having corrected for gross errors, an offset of less than 10 cm can often be obtained and standard deviations are generally well within 15 cm. For hilly and flat terrain densely covered by vegetation, accuracy estimates do not generally fulfil those required by Rijkswaterstaat. However, the use of an adequate strategy for data collection and processing will, to a great extent, improve the accuracy and fidelity of the results. Thus, research should be devoted to the design of appropriate strategies for data collection and processing. Proceedings of ISPRS workshop Sensors and Mapping from Space. University of Hanover, 1999. Available on CD-ROM.
Abstract: While ALS vertical error is typically measured in decimeters the horizontal error can be more than a meter. In steep terrain, this horizontal error necessarily affects the vertical accuracy. Comparing laser data with independent ground truthed data is problematic due to the asymmetric error distribution of laser measurements. The data show only small negative deviations to the terrain-surface (below the terrain-surface), however, relatively big positive deviations due to vegetation or buildings (above the terrain-surface). Application of linear prediction filters should be iterative, otherwise results are strongly influenced by those elevation data lying far above the mid-terrain-level. We show results of linear prediction filtering in forested areas. Accessing LiDAR Data Quality
QualityAssessment_shank.pdf
Shank, M.
TAGIS Unit, W. Virginia Dept. of Environmental Protection. Available from: gis.wvdep.org/tagis/projects/lidar_accuracy.html
Abstract: LiDAR data is validated using stream channel survey data. The comparison is based on coincident points meeting a maximum horizontal separation criteria. Vertical error analysis is highly sensitive to hillslope gradient due to horizontal georeferencing errors. Therefore, ground control points on slopes > 15 degrees are eliminated from the analysis. This improves the apparent agreement between the ALS and survey data but with the consequence that RMSE statistics are only applicable to gentle terrain. A bare earth surface DTM is generated using a TIN model followed by rasterizing by linear interpolation. The advantages of initially using a TIN are, (a) all raw ALS data are used to generate the surface, and (b) the TIN model allows for breakline definition. LiDAR Bibliography:
Last updated: 4/29/2004
Page 23 of 47
Title A Quality Assessment of Airborne Laser Scanner Data
File Name Ahokas_ALSDD2003.pdf
Author(s) Ahokas, E., Kaartinen, H., and J. Hyyppa
Publication and Annotated Abstract Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: ALS data were acquired using two airborne laser scanner systems (Toposys, TopEye) in Southern Finland in September 2002. Toposys flying altitudes were approximately 400-800 meters and 100-550 meters for the TopEye system. Ground survey data were collected for various surfaces, including asphalt, grass, gravel and forest floor. Vertical errors increased as flying height increased. Comparisons based on ALS mean height in the test circle, height of the nearest laser point and interpolated height indicated that mean of differences (reference height – ALS height) in the same flight line were all similar. Data quality also changed across different flight lines. Observation angle also influenced the range of vertical errors, averaging approximately 10-cm..
On the Matching Accuracy of Rasterized Scanning Laser Altimeter Data
behan2000_accuracy.pdf
IAPRS, Vol. XXXIII. Amsterdam, 2000.
Abstract: For certain applications irregularly distributed scanning laser altimeter data needs to be rasterised - such as for use in GIS systems and for creating DEMs. Furthermore least squares matching on a raster grid can enable the measurement of planimetric and height shifts between overlapping strips of laser data. The shifts result from errors in the laser altimeter, primarily the GPS or INS. These shifts form the input into block adjustment to correct for relative and absolute errors. Issues related to generating a regular grid of 2.5D points from the original data with reference to interpolation methodologies, grid size, and quantisation level are also discussed. Surface generation was performed using a TIN and with a horizontal resolution approximately matching the raw horizontal data density. Behan, A., H.G. Maas, G. Vosselman
Steps toward Quality Improvement of Airborne Laser Scanner Data Calibration Procedures of the Image Laser Altimeter and Data Processing
Behan, Avril
thiel.pdf
Thiel, K-H. and A. Wehr
Proceedings of the 26th Annual Conference of the Remote Sensing Society, 12-14 September 2000 Available from: http://www.geo.tudelft.nl/frs/papers/2000/behan2000_quality.pdf Available from: www.nav.uni-stuttgart.de/German/, 1999.
Abstract: The Institute of Navigation of the University of Stuttgart Scanning Laser Altitude and Reflectance Sensor (ScaLARS) work with a continuous wave semiconductor laser diode which is intensity modulated. The amplitude modulation scheme allows for detection of laser light backscattered from Earth surface without background light interference. Post-processing can take therefore use both the 3D elevation data and the intensity data respectively.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 24 of 47
Title Segmentation of TIN-Structured Surface Models
File Name TINsegmentation02.pdf
Author(s) Gorte, Ben.
Publication and Annotated Abstract Joint Symposium on Geospatial Theory, Processing, and Applications. Ottawa, July 8-12, 2002.
Abstract: Segmentation groups laser points into segments that correspond to planar surfaces, such as facets of building roofs or the (flat) terrain between buildings. A segmentation method is presented that was inspired by a raster-based algorithm in the literature but works on triangulated points. It iteratively merges triangles and already formed segments into larger segments. The algorithm is controlled by a single parameter controlling the maximum dissimilarity for adjacent segments such as that merging them is still allowed. The resulting TIN segmentation method is compared with the 3D Hough transform. Analysis and Implementation of a Laser Strip Adjustment Model
Filin_ALSDD2003.pdf
Filin, S.
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: Strip laser adjustment improves the accuracy of laser data and creates a seamless dataset. However, achieving these objectives is difficult. Complex data acquisition systems and numerous error sources make the formulation of a strip adjustment model complex. Difficulties in manually processing the data and limited information consisting only of the laser point coordinates, but not of the system measurements, imply that a strip adjustment is more then just an analytical model. This paper elaborates on both aspects of the problem. We describe the development of a model that is system driven and based on modeling the actual errors in the system. We discuss automatic selection of tie regions in accordance with the proposed model and model implementation using only laser points. Recovery of Systematic Biases in Laser Altimeters Using Natural Surfaces Practical Methods for the Verification of Countrywide Terrain and Surface Models
Artuso_ALSDD2003.pdf
Filin, S.
Available from http://www.geo.tudelft.nl/frs/staff/sagi/pub/Annapolis.pdf
Artuso, R., Bovet, S., and A. Streilein.
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: The Swiss Federal Office of Topography requires that LiDAR DSMs and DTMs meet a height accuracy of 50cm and a mean density of 1 point per m2. Quality control of LiDAR data is accomplished by global and local control techniques.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 25 of 47
Title Quality Improvement of Laser Altimetry DEMs
File Name Elberink_ALSDD2003.pdf
Author(s) Elberink, S.O., Brand, G., and R. Brugelmann
Publication and Annotated Abstract Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: The resolution of the Netherlands national DEM (AHN) is sufficient to meet most land management specifications. However, in peat meadow areas the precision demands are higher because of susceptibly to land subsidence. Mean field heights for agricultural units ranging from 100-700 ha have to be measured with cm-precision. Comparing traditional terrestrial methods (DGPS, tachymetry) with laser altimetry showed that the precision of the AHN does not suffice for mean field height determination in peat meadow areas. However, preliminary block design analysis indicated that adding more GCPs and additional orthogonal strips can improve the elevation accuracy of the DTM. The vertical accuracy improvements were less than expected largely due to strip deformations as a result of long-term GPS/INS positioning errors. The results of this study can be used to help guide flight planning for future ALS projects. Estimating Intrinsic Accuracy of Airborne Laser Data with Local 3DOffsets
Investigating the Spatial Structure of Error in Digital Surface Models Derived from Laser Scanning Data
Bretar_ALSDD2003.pdf
Smith_ALSDD2003.pdf
Bretar, F., PierrotDeseilligny, M., and M. Roux
Smith, S.L., Holland, D.A., and Longley, P.A.
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: A methodology for merging a single laser strip with a photogrammetrically derived DEM is presented. Intra-strip errors are estimated. The approach computes local linear deformations with a tri-dimensional accumulator (in translation space). We show that searching for local discrepancies is equivalent to computing the maximum of the accumulator. Results show a significant improvement when applying local translations to the data. Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: DTM vertical error estimates are typically global estimates for an entire surface. As a result, these estimates ignore the spatially heterogeneous error structure across the surface. We investigate the structure, spatial patterns, and magnitude of these errors by comparing several surface generation routines using different cell sizes.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 26 of 47
Title Estimating relative lidar accuracy information from overlapping flight lines
On the use of pulse reflectance data for laserscanner strip adjustment
File Name
Author(s) Latypov, Damir
Publication and Annotated Abstract ISPRS J. Photogrammetry and Remote Sensing, 56(4):236-245,2002.
Abstract: A pure statistical method for estimating relative lidar accuracy by
Maas, HansGerd
comparing overlapping lidar datasets is described. It has been designed for commercial, fast turnaround, high-volume data production environment and has the following features. It operates on raw data and no building of grid or TIN is required. The method performs accuracy computations on statistical samples that are orders of magnitude larger than ones used in other methods reported in the literature, thus allowing detailed error analysis. Finally, the method provides a lot of relative accuracy information even when no ground control points are available. Software implementation of the method has been written and extensively tested on almost 100 Gb worth of data. Robust and fully automated, it has become an important quality control tool for data processing at TerraPoint. Available from: www.tu-resden.de/fghgipf/forschung/material/publ2001/ Filename: Maas_Annapolis2001.pdf
LiDAR Bibliography:
Last updated: 4/29/2004
Page 27 of 47
5. Commercial LiDAR Title Airborne Laser Scanning: Existing Systems and Firms and Other Resources
File Name alsm_existing.pdf
Author(s) Baltsavias, E.P.
Publication and Annotated Abstract ISPRS J. Photogrammetry and Remote Sensing, 54 (1999): 164-198
Abstract: Currently available ALS systems and commercial LiDAR issues are discussed. A summary of other non-profit and research ASL systems, mainly NASA, and respective links are along with useful web links. Recent developments in the ALS field are reviewed. Enterprise Digital Terrain Modeling
Light Detection and Ranging: Statement of Qualifications to Provide LiDAR Services
edtm.pdf
LiDAR_SOQ_Internet.pdf
Graham, L., Dahman, N., and B. Herman Anon
Z/I Imaging Corp. Huntsville, AL. October 30, 2001
Abstract: We present a holistic, ‘enterprise’ approach to processing elevation data. The tools are aimed at increasing the production throughput and, as a result, the profit of companies engaged in processing elevation data. www.merrick.com
Abstract: ALS technology is discussed through the experiences of Merrick.
Raw
ALS data acquisition through to bare-earth surface generation using Merrick’s proprietary MARS software, is discussed.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 28 of 47
6. Forestry Applications Title
File Name
Author(s) Næsset, Erik
Determination of mean tree height of forest stands using airborne laser scanner data
Publication and Annotated Abstract ISPRS J. Photogrammetry and Remote Sensing, 52(2):49-56, 1997.
Abstract: The mean tree height of forest stands is a crucial stand characteristic in forest planning. Currently, the mean tree height is determined by field measurements or by photogrammetric measurements utilizing aerial photographs. In this study, mean tree height of 36 test stands is derived from tree canopy heights measured by means of an airborne laser scanner. On the average the laser recorded 505–1070 canopy heights per stand. First, the laser mean height is computed as the arithmetic mean of the canopy heights within each stand. The laser mean height underestimates the ground truth mean height by 4.1–5.5 m. Second, a weighted mean of the laser canopy heights is computed. The individual height values are used as weights. The weighted mean height underestimates the true height by 2.1–3.6 m. Finally, the laser mean height is computed as the arithmetic mean of the largest laser values within square grid cells with cell sizes of 15–30 m. The bias of the laser estimates is in the range 0.4 m to 1.9 m. The standard deviation for differences between the laser mean heights and the ground truth mean height is 1.1–1.6 m.
Three-Dimensional Analysis of Forest Structure and Terrain using LiDAR Technology
Interpolation of High Quality Ground Models from Laser Scanner Data in Forested Areas
3DanalysisForestLidar.pdf
Wulder, M., StOnge, B., and P. Treitz
GEOIDE Calgary 2000, From Ideas to Innovation – Geomatics for a New Millennium, May 25-26, 2000, Calgary, AB.
Abstract: The GEOIDE project is evaluating algorithms for estimating forest structure and biophysical attributes. The three main thrusts of the project are: (1) measuring forest structural and biophysical variables, (2) measuring forest geometry, and (3) simulation of space-borne LiDAR data from airborne LiDAR. InterpModels_forests.pdf
Pfeifer, N., et al.
International Archives of Photogrammetry and Remote Sensing. WG III/5 and WG III/2. 9-11 Nov. 1999, Vol. 32, Part 3W14. La Jolla, CA, USA. pp. 31-36.
Abstract: We describe an iterative linear prediction filtering approach which is used to extract bare-earth DTMs. The approach is compared to other filtering strategies currently in use. Accuracy analysis is shown for two separate ALS datasets.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 29 of 47
Title A Multiple Resource Inventory of Delaware Using Airborne Laser Data
File Name Nelson et al LiDAR bioscience 2003.pdf
Author(s) Nelson, R., et al.
Publication and Annotated Abstract Bioscience, Volume 53(10): 981-992. 2003.
Abstract: Estimates of forest wood volume and biomass, and estimates of the surface area of asphalt, concrete, roof, and open water, were generated from a single set of airborne laser-profiling data acquired during the summer of 2000. Estimates of aboveground dry biomass for different types of land cover in each county or state were converted directly to estimates of standing carbon. A portable, inexpensive laser that measures forest canopy height and canopy cover was also used to identify and map mature forest stands that might support the Delmarva fox squirrel (Sciurus niger cinereus), an endangered species. Merchantable volume estimates were within 24 percent of US Forest Service estimates at the county level and within 15 percent statewide. Various types of impervious surfaces (roofs, concrete, asphalt) and open water were tallied along the flight lines to estimate areal coverage statewide, by land cover and county. Laser estimates of impervious surface area were within 28 percent of satellitebased estimates at the county level and within 3 percent at the state level. Laser estimates of open water were within 7 percent of photointerpreted geographic information system (GIS) estimates at the county level and within 3 percent of the GIS at the state level.
Canopy height models and airborne lasers to estimate forest biomass: two problems Modeling LiDAR Returns from Forest Canopies
Canopy_waveformreturns.pdf
Nelson, R., et al.
Int. J. Remote Sensing, 21(11): 2153-2162, 2000.
Guoqing, S., and K.J. Ranson
IEEE Trans. Geosci. and Remote Sensing, 38(6). November, 2002.
Abstract: We describe a 3D model which is being used to investigate the relationship between canopy structure and the lidar return waveform. Detailed field measurements and forest growth model simulations of forest stands were used to parameterize the model. Crown shape determines the vertical distribution of plant material in the model and the corresponding lidar waveforms. Preliminary comparisons of averaged waveforms from an airborne lidar and model simulations shows that the shape of the measured waveform was more similar to simulations using an ellipsoid or hemi-ellipsoid shape. The observed slower decay of the airborne lidar waveforms than the simulated waveforms may indicate the existence of the understories and may also suggest that higher order scattering from the upper canopy may contribute to the lidar signals. The lidar waveforms from stands simulated from a forest growth model show the dependence of the waveform on stand structure.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 30 of 47
Title The Laser Vegetation Imaging Sensor: a MediumAltitude, Digitization-only, Airborne Laser Altimeter for Mapping Vegetation and Topography
Quantifying canopy height underestimation by laser pulse penetration in smallfootprint airborne laser scanning data. Proceedings of the ScandLaser Scientific Workshop on Airborne Laser Scanning of Forests
Accuracy of a HighResolution LiDAR Terrain Model under a Conifer Forest Canopy
LiDAR Bibliography:
File Name
Author(s)
waveformlidar.pdf
Blair, J.B, Rabine, D.L., and M. A. Holton
Publication and Annotated Abstract Journal Photogrammetry and Remote Sensing, Volume 54: 115-122. 1999.
Abstract: The Laser Vegetation Imaging Sensor (LVIS) is an airborne, scanning laser altimeter, designed and operated at NASA’s Goddard Space Flight Center. LVIS operates at flying heights of up to 10 km and with 1000meter swatch widths, nominally with 25-m wide footprints. The entire LiDAR waveform is digitised allowing for unambiguous determination of range and return pulse structure. LVIS has been used to acquire sub-canopy and vegetation height and structure data for several forested areas in the US and Central America. LVIS is the airborne simulator for the Vegetation Canopy Lidar (VCL) mission, a NASA remote sensing satellite due for launch in 2000.
V5_second_edition_031020.pdf
D.L.A Gaveau and R.A. Hill
Canadian Journal of Remote Sensing, 29:650-657, 2003.
Edited by J. Hyyppa et al.
Scandlaser Scientific Workshop, Umea, Sweden. September 3-4, 2003. 275pp
Abstract: Includes the entire proceedings of the 2003 Scandinavian Workshop on Laser Scanning of Forest Resources. Sessions included: Session 1 History and State-of-Art in Scandinavia Session 2 System parameters, models, and algorithms Session 3 Ecological applications Session 4 Change studies Session 5 Forest inventory applications Session 6 Single tree based methods Poster Session Practical workshop paper. m03-022.pdf
Reutebuch, S.E. et al
Canadian Journal of Remote Sensing, Vol. 29(5): 527-535, 2003.
Abstract: A high-resolution digital terrain model (DTM) was produced from ALS data for a 500-ha study area containing various surface cover types, including bare ground and dense, conifer forest. Survey data for for 347 GCPs distributed under a range of canopy covers were used to test DTM accuracy. The mean DTM error was 0.22 ± 0.24 m (mean ± SD). DTM elevation errors for four tree canopy cover classes were: clearcut 0.16 ± 0.23 m, heavily thinned 0.18 ± 0.14 m, lightly thinned 0.18 ± 0.18 m, and uncut 0.31 ± 0.29 m. DTM errors show only a slight increase with canopy density.
Last updated: 4/29/2004
Page 31 of 47
Title
File Name
Use of Large-Footprint Scanning Airborne LiDAR to Estimate Forest Stand Characteristics in the Western Cascades of Oregon
Alsm_canopy.pdf
Processing of Laser Scanning Data for Wooded Areas
LaserForWoodedAreas.pdf
Author(s) J.E. Means and 7 other authors
Publication and Annotated Abstract Remote Sensing of Environment, 67: 298-308. 1998.
Abstract: NASA’s SLICER (Scanning LiDAR Imager of Canopies by Echo Recovery), scans a swath width of five 10-m diameter footprints along the aircraft’s flightpath, and captures the power of the reflected laser pulse as a function of height from the top of the canopy to the ground. Ground measurements of forest stand structure were collected for 26 plots in the Pacific Northwest. Height, basal area, total biomass, and leaf biomass as estimated from field data could be pre-indicted from SLICER-derived metrics with R2 values of 0.95, 0.96, 0.96, and 0.84, respectively. These relationships were strong up to a height of 52 m, basal area of 132 m2/ha and total biomass of 1300 Mg/ha. Kraus, K., and W. Rieger
Downloaded from: http://www.ifp.unistuttgart.de/publications/phowo99/phowo99.htm
Abstract: We present a new ALS filtering technique which assesses the skewness the error distribution function and assigns small weights to those points with large positive errors. This results in an approximate classification of bare-earth and non bare-earth (chiefly vegetation) ALS points. The approach uses both first and last return pulses and is implemented for a grid-based DTM. Methods are presented for extraction of forest stand information, including mean stand height, stand density, and tree species. Airborne Laser Measurements of Rangeland Canopy Cover and Distribution
LiDAR Bibliography:
RangelandCanopyCover.pdf
J.C. Ritchie and 4 other authors
J. Range Management, 45: 189-193. 1992
Abstract: Studies were made for 2 rangeland areas in south Texas to measure canopy cover and distribution using an ALS. Compared to ground surveyed canopy data, laser measurements ranged from 1 to 89% and were correlated significantly (P = 0.89) with ground measurements (1 to 88%). Topography, vegetation height, and spatial distribution of canopy cover were also measured with the laser profiler. Accurate and timely data on the amount and distribution of plant cover are valuable for understanding vegetation characteristics, improving estimates of infiltration, erosion, and evapotranspiration for rangeland areas, and making decisions for managing rangeland vegetation.
Last updated: 4/29/2004
Page 32 of 47
Title
File Name
Author(s)
An Evaluation of LiDAR and IFSAR Derived Digital Elevation Models in Leaf-on Conditions with USGS Level 1 and Level 2 DEMs
remo-sens-envi-84_2003.pdf
Hodgson, M.E., Jensen, J.R. and 3 others.
Elevation Accuracy of Laser Scanning-Derived Digital Terrain and Target Models in Forest Environment
TargetModelsForestEnvironment.pdf
First Experience in the Application of Laser Scanner Data for the Assessment of Vertical and Horizontal Forest Structures
VertHorizForestStructure.pdf
LiDAR Bibliography:
Hyyppa, J., Pyysalo, U., Hyyppa, H., and A. Samberg
Friedlaender, H., and B. Koch.
Publication and Annotated Abstract Remote Sensing of Environment, 84: 295-308 (2003)
Abstract: Four approaches (LiDAR, IFSAR, Gestalt Photomapper (GPM), and USGS contour-to-grid) for DEM generation in floodprone watersheds in North Carolina, were compared. Ground survey data (1,470 points) for five different land cover classes were used to assess DEM accuracies. One unique aspect of this study was the LiDAR and IFSAR data were collected during leafon conditions. Analyses of absolute elevation accuracy and terrain slope were conducted. The LiDAR and contour-to-grid derived DEMs exhibited the highest overall absolute elevation accuracies. Elevation accuracy was found to vary with land cover categories, decreasing with increasing slopes—but only for the scrub/shrub land cover category. Large slope errors were observed for all 4 methods. 4th EARSel Workshop on LiDAR Remote Sensing of Land and Sea. 2000.
Abstract: We examine ALS data accuracy with the objective of building tree height DTMs. Special emphasis is placed on optimizing the process of ground surface data selection for the creation of the DTM. A novel DTM algorithm is depicted in detail. The algorithm follows a 5-tier strategy: 1) Calculation of the original reference surface, 2) Removal of the vegetation from the reference surface, 3) Classification of the original point cloud, 4) Calculation of the DTM based on the classified ground hits, and 5) Interpolation of the missing points. A standard error of 15 cm was obtained for flat forest areas, increasing with steeper slopes to approximately 40 cm for 40% slopes. The average standard error for forest areas was about 22 cm. The laser-derived DTM of the forest road deviated from the true height by 8.5 cm only. Optimum DTM performance was obtained by averaging the ground hits which were located at a maximum, of 60 cm above the minimum terrain values. We also demonstrate that individual tree heights in the dominating stands can be obtained with RMSE’s less than 1meterj. IAPRS, Vol. XXXIII. Amsterdam, 2000.
Abstract: Several approaches for the assessment of horizontal and vertical forest stand structures are presented. First results are presented and discussed.
Last updated: 4/29/2004
Page 33 of 47
7. Urban Applications Title Extraction of Buildings and Trees in Urban Environments
File Name ExtractBuildTreeUrbanEnv.pdf
Author(s) Haala, N., and C. Brenner
Publication and Annotated Abstract ISPRS J. Photogrammetry and Remote Sensing, 54: 130-137, 1999.
Abstract: We present two methods for data collection in urban environments. The first integrates color-infrared imagery and ALS data to generate a DSM. A bare-earth DTM is generated after filtering with a morphological filter. The output from subtracting the DSM from DTM is used as input to an unsupervised classification algorithm. The algorithm detects similar clusters based on minimum distance criterion and can be used to extract simple 2.5D urban models. We show a more sophisticated approach to urban model generation which uses laser data and 2D ground plan information to obtain 3D reconstructions of buildings. Haala, N., and C. Brenner
Interpretation of Urban Surface Models Using 2D Building Information
Comp. Vis. Image Inderstanding, 72(2): 204-214, 1998.
Abstract: In 3D building reconstruction the interpretation process can be simplified if digital surface models (DSM), which can either be derived from stereo matching of aerial images or be directly measured by scanning laser systems, are used in addition to or instead of image data. Images contain much information, but the resulting complexity causes enormous problems for an automatic interpretation of this data type. Since the information of a DSM is restricted to surface geometry its interpretation is simplified by the absence of unnecessary details. Nevertheless, due to insufficient spatial resolution and quality of the DSM, especially for these applications, optimal results can only be achieved by the use of additional data sources. Within the approach presented in this paper the segmentation of planar surfaces from the DSM is supported by existing ground plans. This 2D building information is also used to derive hypotheses on the possible roof shapes in order to obtain a 3D boundary representation based on the segmented planes.
Interpolation of LiDAR Data and Automatic Building Extraction
LidarInterp_Morgan.pdf
Morgan, M., and A. Habib
Proceedings of the ASPRS/ASCM Conference, Washington, D.C. (April 19-26, 2002).
Abstract: We describe how improvements can be made to raw LiDAR data interpolation by extracting breaklines. A procedure for building detection and extraction from the DSM is developed. A region growing algorithm based on least-squares adjustment of laser data connected by a TIN to extract the building facades is introduced. The variance is used to estimate the accuracy and the validity of the extracted parameters. Adjacent planar facades are used to extract the 3D breaklines which can be stored in vector format or can be used to separate the uncorrected segments to guide any subsequent interpolation processes or building extraction. A morphological filter is used to separate bare-earth from non-terrain areas. The resulting classification combined with the extracted planar facades and breaklines, will be used for extracting building geometries. LiDAR Bibliography:
Last updated: 4/29/2004
Page 34 of 47
Title Building Extraction from LiDAR Data
File Name Building Extraction from LiDAR Data.pdf
Author(s) Rottensteiner, F., and C. Briese
Publication and Annotated Abstract Available from: http://www.ipf.tuwien.ac.at/research/fr_buildings_lidar/buildings_lidar.htm
Institute of Photogrammetry and Remote Sensing, Vienna University of Technology. 2002.
Abstract: An hierarchical application of robust interpolation using a skew error distribution function is used to separate ALS-acquired bare-earth from non bareearth objects. Building features are separated from other non bare-earth objects by analysing the height differences of the DSM, the DTM, and original raw data point cloud. The result is a ‘building mask’ or training area, which is then used to classify other urban areas using a curvature-based segmentation technique. Preliminary results are shown for a Vienna test site.. 3D Resampling for Airborne Laser Data of Urban Areas
3Dresampling.pdf
Two Algorithms for Extracting Building Models from Raw Laser Altimetry Data
Buildingextractionalgorithms.pdf
Integration of Laser-Derived DSMs and Matched Edges for Generating an Accurate Surface Model
Laser_DSMintegration.pdf
LiDAR Bibliography:
Zinger, S., Nikolova, M., Roux, M., and H. Maitre
Maas, H-G., and G. Vosselman
ISPRS Commission III, Symposium 2002, September 9 - 13, Graz, Austria
Abstract: We consider DSM interpolation methods using raw ALS data. TINbased linear interpolation, TIN-based nearest neighbor interpolation and kriging are compared. We propose an energy (or cost function) minimization approach which avoids some of the drawbacks of these approaches. An adaptive energy function is developed which is parameterized specifically for urban land cover types. ALS data for Brussels are used to test and compare the new approach. ISPRS J. Photogrammetry and Remote Sensing, 54: 153-163, 1999.
Abstract: Two new techniques for the determination of building models from laser altimetry data are presented, both of which use all the original laser scanner data. Ground plan information may be used, but is not required. Closed solutions for the determination of the parameters of a standard gable roof type building model based on invariant moments of 2.5D-point clouds are shown. In addition, the analysis of deviations between point cloud and model allows for modelling asymmetries such as dorms on a gable roof. The techniques were applied to a FLIMAP laser scanner dataset with a point density of > 5 points/m2. ANOVA analysis indicates a precision in the range of 0.1–0.2 m. McIntosh, K., and A. Krupnik
ISPRS J. Photogrammetry and Remote Sensing, 56: 167-176, 2002.
Abstract: Orthophotos and ALS data are used to generate accurate surface models of urban environments. The approach, which uses TIN interpolation with defined breaklines, better represents surface discontinuities, particularly building walls. Our results show that the surface accuracy is improved by 49% and 15% for two test sites (RMSE’s of 0.52m and 0.53m respectively).
Last updated: 4/29/2004
Page 35 of 47
Title Automatic Detection of Buildings from Laser Scanner Data for Map Updating
File Name
Author(s)
Matikainen_ALSDD2003.pdf
Matikainen, L., Hyyppa, J., and H. Hyyppa.
Publication and Annotated Abstract Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: Approaches to the automatic extraction of laser scanner data for building detection and map updating are presented. A DSM, DTM, and an intensity image were generated from raw ALS data. The DSM was segmented into `building´, `tree´ or `ground surface´ land classes using a region-based segmentation method. Height differences between the DSM and DTM, textural characteristics of the DSM and intensity image, and segment shape, were used in the classification. Comparison with a new building map shows that about 80% of all buildings were correctly detected. Buildings larger than 200-m2 were correctly identified 90% of the time.
A New Method for Building Extraction in Urban Areas From High-Resolution LiDAR Data
Rottensteiner_Briese.pdf
Rottensteiner, F., and Ch. Briese
Institute of Photogrammetry and Remote Sensing, Vienna University of Technology. Downloaded from: www.ipf.tuwien.ac.at/research/
Abstract: A hierarchic application of robust interpolation using a skew error distribution function is used to extract 3D building models. The approach analyses the height differences between a DSM, the original raw LiDAR data and a bareearth DTM. Polyhedral building models are created in candidate regions using a bottom-up procedure by applying curvature-based segmentation techniques..
On the Quality of Object Classification and Automated Building Modelling Based on Laser Scanning Data
Voegtle_ALSDD2003.pdf
Voegtle, T., and E. Steinle
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: We describe a method for automatic bare-earth data extraction that applies a region growing algorithm and which segments 3D objects across the terrain surface. Features are extracted using fuzzy logic statistics. Accuracy analysis applies classification rates which are compiled in a confusion matrix. For geometric modelling of buildings, a generic approach is used whereby buildings are approximated by (oblique) planes. The intersection of neighbouring planes leads to building edges and corners. Extracting objects from digital terrain models
LiDAR Bibliography:
Eckstein, W., Munkelt, O.
Remote Sensing and Reconstruction for Three-Dimensional Objects and Scenes, Proc. Spie 2572, pages 222-231, 1995.
Last updated: 4/29/2004
Page 36 of 47
Title Hierarchical Bayesian nets for extraction using dense digital surface models
File Name
Author(s) Brunn, A., Weidner, U.
Publication and Annotated Abstract ISPRS J. Photogrammetry and Remote Sensing, 53(5): 296-307, 1998.
Abstract: During the last years an increasing demand for 3D data of urban scenes can be recognized. Techniques for automatic acquisition of buildings are needed to satisfy this demand in an economic way. This paper describes an approach for building extraction using digital surface models (DSM) as input data. The first task is the detection of areas within the DSM which describe buildings. The second task is the reconstruction of geometric building descriptions. In this paper we focus on new extensions of our approach. The first extension is the detection of buildings using two alternative classification schemes: a binary or a statistical classification based on Bayesian nets, both using local geometric properties. The second extension is the extraction of roof structures as a first step towards the reconstruction of polyhedral building descriptions.
Application of snakes and dynamic programming optimisation technique in modeling of buildings in informal settlement areas
Rüther, H., Martine, H. M., Mtalo, E. G.
Closed solutions for the determination of parametric building models from invariant moments of airborne laserscanner data
Maas, H.-G.
Wavelet and scalespace theory in segmentation of airborne laser scanner data
Vu, T. T., Tokunaga, M.
Extraction of Digital Elevation Models for Airborne Laser Terrain Mapping Data
Neuenschwander, A., Crawford, M., Weed, C., Guitierrez, R.
LiDAR Bibliography:
ISPRS J. Photogrammetry and Remote Sensing, 56(4): 269-282, 2002.
Abstract: This paper presents a novel approach to semiautomatic building extraction in informal settlement areas from aerial photographs. The proposed approach uses a strategy of delineating buildings by optimising their approximate building contour position. Approximate building contours are derived automatically by locating elevation blobs in digital surface models. Building extraction is then effected by means of the snakes algorithm and the dynamic programming optimisation technique. With dynamic programming, the building contour optimisation problem is realized through a discrete multistage process and solved by the "time-delayed" algorithm, as developed in this work. The proposed building extraction approach is a semiautomatic process, with user-controlled operations linking fully automated subprocesses. Inputs into the proposed building extraction system are ortho-images and digital surface models, the latter being generated through image matching techniques.
Presented at ISPRS Conference ‘Automatic Extraction of GIS Objects from Digital Imagery,’ München, Germany, 8-10 September, 1999, (IAPRS Vol. 32, Part 3-2W5, pp. 193-199) Available from: http://www.tudresden.de/fghgipf/forschung/material/publ_maas/Maas_Munich99.pdf Presented at 22nd Asian Conference on Remote Sensing 5-9 November 2001, Singapore Available from: http://www.crisp.nus.edu.sg/~acrs2001/pdf/037VU.PDF
Proceedings to the 2000 International Geoscience and Remote Sensing Symposium, Honolulu, Hawaii, July 24 - 28, pp. 2305 - 2309, 2000.
Last updated: 4/29/2004
Page 37 of 47
Title Extracting urban features from LiDAR digital surface models
File Name
Author(s) Priestnall, G., Jaafar, J., Duncan, A.
Reconstructing 3D Buildings from LIDAR data Heuristic Filtering and 3D Feature Extraction from LIDAR data Automated Road Extraction from LiDAR Data Urban Object Recognition Using Airborne Laser Elevation Image and Aerial Image
Elaksher, A. F., Bethel, J. S. Alharthy, A., Bethel, J.
Towards Fully Automatic Generation of City Models
Brenner, Claus
LiDAR Bibliography:
Alharthy, A., Bethel, J. Fujii, K., Arikawa, T.
Publication and Annotated Abstract Computers, Environment, and Urban Systems, 24(2): 65-78, 2000.
Abstract: The use of airborne Light Detection And Ranging (LiDAR) technology offers rapid high resolution capture of surface elevation data suitable for a large range of applications. The representation of both the ground surface and the features on that surface necessitates the removal of these surface features if a ground surface Digital Elevation Model (DEM) product is to be produced. This paper examines methods for extracting surface features from a Digital Surface Model (DSM) produced by LiDAR. It is argued that for some applications the extracted surface feature layer can be of almost equal importance to the DEM. The example of flood inundation modelling is used to illustrate how a DEM and a surface roughness layer could be extracted from the original DSM. The potential for refining surface roughness estimates by classifying extracted surface features using both topographic and spectral characteristics is considered using an Artificial Neural Network to discriminate between buildings and trees. Available from: http://www.isprs.org/commission3/proceedings/papers/paper102.pdf Available from: http://www.isprs.org/commission3/proceedings/papers/paper061.pdf
EOM, 2003 IEEE Trans. Geo. Remote Sens., 40(10):2234-2240, 2002.
Abstract:
Creating three-dimensional (3-D) models of real urban objects is an important goal in a wide variety of applications. This paper describes a method that utilizes airborne laser elevation images and aerial images for the 3-D reconstruction of urban objects. Our modeling approach uses the vertical geometric pattern analysis of elevation images. These patterns correspond to object contours and, thus, enable the extraction of the object. In addition, to provide realistic textured details, textures are cut from aerial images and mapped onto 3-D models. Our texture-mapping approach can avoid geometry mismatching and enable the automatic registration to determine the most reliable correspondence between projected outlines of 3-D models and contours of real objects shown in aerial images. Edge pairs, which are matched with projected outlines, are detected from aerial images. In order to minimize mismatching, we apply the voting technique based on the generalized Hough transform. Experimental results show that 3-D reconstruction of urban objects is generally successful. Available from: http://www.ifp.uni-stuttgart.de/publications/2000/Brenner_amsterdam.pdf
Last updated: 4/29/2004
Page 38 of 47
Title
File Name
Author(s)
The Use of Anisotropic Height Texture Measures for the Segmentation of Airborne Laser Scanner Data The Potential of Height Texture Measures for the Segmentation of Airborne Laserscanner Data
Elberink, S. O., Maas, H.-G.
Building Footprint Extraction and 3-D Reconstruction from LIDAR Data
Haithcoat, T. L., Song, W., Hipple, J. D.
LiDAR Bibliography:
Publication and Annotated Abstract IAPRS, Vol. XXXIII, Amsterdam, 2000. Available from: http://www.geo.tudelft.nl/frs/papers/2000/sanderoe.pdf
Maas, Hans-Gerd
Presented at the 4th International Airborne Remote Sensing Conference Ottawa, 21-24 June, 1999. Available from: http://www.geo.tudelft.nl/frs/papers/1999/Maasthepotential.pdf
IEEE/ISPRS Joint Workshop on Remote Sensing and Data Fusion over Urban Areas.
Last updated: 4/29/2004
Page 39 of 47
8. Other Applications Title
File Name
Using Airborne Laser Altimetry and GIS to Assess Scale-Induced Radiation Loading Errors in a Glacierised Basin
Chasmer.pdf
Fractal Modeling of Airborne Laser Altimetry Data
fractal_als_modelling.pdf
LiDAR Bibliography:
Author(s) Chasmer, L., and C. Hopkinson
Publication and Annotated Abstract th
58 Eastern Snow Conference. Ottawa, Ontario, Canada, 2001
Abstract: We compare the results of a short-wave radiation model applied over a 25-metre and a 2.5-meter DEM of Peyto Glacier in the Canadian Rocky Mountains. Results show that radiation load is primarily influenced by terrain slope which reduces radiation load upon a surface by between 1 and 6 W/m2 as computed from the higher resolution LiDAR DTM. Pachepsky, Y.A., Ritchie, J.C., and D. Gimenez
Last updated: 4/29/2004
REMOTE SENS. ENVIRON. 61:150-161 (1997)
Abstract: A k-borne laser altimetry is a remote sensing technique that can provide high resolution data on the roughness of the landscape both for estimating water balance components and for distinguishing between landscapes. Models of the scale-dependent roughness are needed to find scales most appropriate for these purposes. Our objectives were to apply fractal scaling to high-resolution profiling laser altimetry data and to determine fractal parameters for differentiating land cover. Data were collected at the USDA-ARS Experimental Range in New Mexico over grass-dominated and shrub-dominated sites along four transects at each site. Scale-dependent RMS roughness and data power spectrums were computed from 100,000 data points (~2 km) from each transect. A linearity measure and piecewise linear approximation were applied to find intervals of the fractal scaling. The RMS roughness data had two intervals of self-affine fractal scaling on grass transects and four such intervals on shrub transects, Reduction in the number of data points did not lead to a decrease in roughness but caused a smoothing dependency of fractal dimension on scale. Ten- and hundred-meter scales were appropriate for distinguishing between grass and shrub transects on the basis of fractal dimensions..
Page 40 of 47
Title Fault Scarp Detection Beneath Dense Vegetation Cover: Airborne Laser Mapping of the Seattle Fault Zone, Bainbridge Island, Washington State
Wavelet-Based Analysis for Object Separation from Laser Altimetry Data
File Name Harding2000.pdf
Author(s) Harding, D.J., and G.S. Berghoff
Publication and Annotated Abstract Proceedings of the Am. Soc. Of Photogrammetry and Remote Sensing Annual Conference, Washington, D.C. May, 2000
Abstract: The emergence of a commercial airborne laser mapping industry, inspired by NASA technology research and development, is paying major dividends in an assessment of earthquake hazards in the Puget Lowland of Washington State. Geophysical observations and historical seismicity indicate the presence of active uppercrustal faults in the Puget Lowland, placing the major population centers of Seattle and Tacoma at significant risk. However, until recently the surface trace of these faults had never been identified, neither on the ground nor from remote sensing, due to cover by the dense vegetation of the Pacific Northwest temperate rainforests and extremely thick Pleistocene glacial deposits. A pilot lidar mapping project of Bainbridge Island in the Puget Sound, contracted by the Kitsap Public Utility District (KPUD) and conducted by Airborne Laser Mapping in late 1996, spectacularly revealed geomorphic features associated with fault strands within the Seattle fault zone. The success of this pilot study has inspired the formation of a consortium of federal and local organizations to extend this work to a 2350 square kilometer region of the Puget Lowland. Wavelet_application.pdf
Amgaa, Tsolmon
Unpublished PhD dissertation, International Institute for GeoInformation Science and Earth Observation, Enschede, Netherlands. February, 2003
Abstract: I investigate the potential for applying wavelet-based analyses for object extraction from laser scanning data. Several wavelet transforms were investigated, including, standard pyramidal discrete wavelet transform, translation invariant wavelet transform, fast lifted wavelet transform, and continuous wavelet transform. Experiments were carried out on complex urban scene. The approach applies wavelet transforms to decompose the original image into detailed and approximate sub-images, perform object and feature extraction, and then inversely transform the sub-images back into single image. Tests on edge extraction, 3D edges, reconstruction of objects by 3D edges locally, and inverting the wavelet coefficients were performed.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 41 of 47
Title Modeling Airborne Laser Scanning Data for the Spatial Generation of Critical Forest Parameters in Fire Behavior Modeling
Suitability of Laser Data for Deriving Geographical Information: A Case Study in the Context of Management of Fluvial Zones
File Name
Author(s)
FireModeling.pdf
Riano, D., Meier, E., Allgower, B., Chuvieco, E., and S.L. Ustin
Lidar_fluvialextraction.pdf
Gomes Pereira, L.M., and R.J. Wicherson
Publication and Annotated Abstract Remote Sensing of Environment, Volume 86: 177-186. 2003
Abstract: A model for the automatic extraction of forest information which can be used to provide input for fire behaviour models, is presented. A DSM and DTM were generated from raw ALS data. Data were defined as bare-earth if heights were less than 0.6 m. A cluster analysis was used to discriminate crown base height, allowing identification of both tree and understory canopy heights. Tree height was defined as the 99th percentile of the tree crown height group, while crown base height was the 1- percentile of the tree crown height group. Tree cover was estimated from the fraction of total tree laser hits relative to the total number of laser hits. Surface canopy height was computed as the 99th percentile of the surface canopy group. ISPRS J. Photogrammetry and Remote Sensing, 54: 153-163, 1999.
Abstract: Flood management in the Netherlands requires detailed terrain data which has traditionally been obtained by photogrammetric techniques. These techniques are expensive and time-consuming, requiring at least 4-years to fully capture the necessary terrain information. ALS offers a cheaper and faster alternative. We assess the feasibility of using laser data to generate a hydrodynamic model which will be used to determine high water levels and their impact on earthworks. The ALS data appear to accurately characterize fluvial areas and are of sufficient quality to be used for national flood management efforts.
Suitability of laser data for DTM generation: a case study in the context of road planning and design
Gomes Pereira, L.M., Janssen, L. L. F.
ISPRS J. Photogrammetry and Remote Sensing, 54(4): 244-253
Abstract: Laser range data acquired from a helicopter are evaluated in terms of the information that can be derived from them and the accuracy. The objective is to study the suitability of laser data to generate a DSM for road planning and design in The Netherlands. The conclusion is that high-density laser measurements allow the reconstruction of the terrain relief with the required accuracy. Nonetheless, they do not allow the extraction of all the information required, particularly semantic information. Thus, the combination of laser data with existing information is a prerequisite. This process of combining laser data with existing geographic information is not trivial. The rate of success depends much on the quality of the individual datasets and the method used to combine them. This problem appears in a much broader context, that of spatial data fusion, and should be the object of future research.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 42 of 47
Title Landscape Modelling Using Integrated Airborne Multispectral and Laser Scanning Data
Landscape visualization: rendering a virtual reality simulation from airborne laser altimetry and multispectral scanning data Assigning the 3rd Dimension to Roads by Use of Laser Scanning Data
File Name Lidar_landcover.pdf
Isprs_hatger_2002.pdf
Author(s) Hill, R.A., Smith,G.M., Fuller, R.M, and N. Veitch
Publication and Annotated Abstract Int. J. Remote Sensing, 23(11): 2327-2334, 2002.
Abstract: Integrating multi-spectral and elevation data from airborne sensors (CASI and ALTM) and adopting a parcel-based approach significantly improves landscape modelling efforts. We describe an approach to data integration, parcel classification, knowledge-based correction, and the derivation of landscape objects. For a 1 km2 study area, a 14 land-cover class vector dataset was generated in which the parcels relate to landscape objects and contain information on structure and terrain. At a 1m horizontal resolution, an 88% correspondence was achieved between ALS land-cover data and field surveyed data acquired in 2000.
Hill, R.A., N. Veitch
Int. J. Remote Sensing, 23(17): 3307-3309, 2002.
Hatger, C.
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol. 34, Part XXX. 2002
Abstract: An approach to identify road attributes, specifically outline, height, longitudinal and traversal slope, width and curvature objects, using LiDAR data is presented.
3D Reconstruction of Roads and Trees for City Modelling
Vosselman_ALSDD2003.pdf
Vosselman, G.
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: We present a method for street surface reconstruction using airborne laser data and cadastral map information. Locations of trees are also extracted from the laser data.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 43 of 47
Title
File Name
Author(s)
Publication and Annotated Abstract
Image processing of Airborne scanning laser altimetry data for improved river flood modeling
Cobby, D. M., Mason, D. C., Davenport, I. J.
Quantifying Riparian Vegetation and Stream Bank Form through the Use of Airborne Laser Scanning and Digital Video Data Quantifying woodland structure and habitat quality for birds using airborne laser scanning Improving bird Population models using airborne remote sensing
Witte, C., Dowling, R., Weller, D., Denham, R., Rowland. T.
IEEE, 2001.
Hinsley, S. A., Hill, R. A., Gaveau, D. L. A., Bellamy, P. E.
Functional Ecology, 16: 851-857, 2002.
Davenport, I. J., Bradbury, R. B., Anderson, G. Q. A., Hayman, G. R. F., Krebs, J. R., Mason, D. C., Wilson, J. D., Veck, N. J.
Int. J. Remote Sensing, 21(13): 2705-2717, 2000.
LiDAR Bibliography:
ISPRS J. Photogrammetry and Remote Sensing, 56(2): 121-138, 2001.
Abstract: This paper describes a range image segmentation system for data from a LiDAR measuring either time of last significant return, or measuring time of both first and last returns. We focus on the application of the segmenter to improving the data required by 2D hydraulic flood models, i.e. maps of topographic height which provide model bathymetry, and vegetation height, which could be converted to distributed floodplain friction coefficients. In addition, the location of river channels and a suitable height contour are used to define the extent of the model domain. An advantage of segmentation is that it allows different topographic and vegetation height extraction algorithms to be used in regions of different cover type. LiDAR data for a reach of the River Severn, UK, is presented. Short vegetation heights (grass and cereal crops) are predicted with a rms error of 14 cm. The topography underlying such cover differs from manually measured spot heights by 17 cm (rms error). The topographic accuracy decreases in the presence of a densely wooded slope. Errors in the vegetation height map, apparent at the overlap regions of adjacent swaths, are reduced by the removal of heights measured at large scan angles.
Last updated: 4/29/2004
Page 44 of 47
Title
File Name
Woolard, J. W., Colby, J. D.
Spatial characterization, resolution, and volumetric change of coastal dunes using airborne LIDAR: Cape Hatteras, North Carolina
Some examples of European activities in airborne laser techniques and an application in glaciology
Author(s)
Publication and Annotated Abstract Geomorphology,48: 269-287, 2002.
Abstract: The technological advancement in topographic mapping known as airborne Light Detection and Ranging (LIDAR) allows researchers to gather highly accurate and densely sampled coastal elevation data at a rapid rate. The problem is to determine the optimal resolutions at which to represent coastal dunes for volumetric change analysis. This study uses digital elevation models (DEM) generated from LIDAR data and spatial statistics to better understand dune characterization at a series of spatial resolutions glaciology_application.pdf
Faveya,E, Wehrb, A., Geigera, A., and H.G. Kahlea
Journal of Geodynamics 34 (2002) 347–355
Abstract: Airborne Laser Altimetry (ALA) has experienced a rapid increase in popularity as a method serving a wide range of applications in Remote Sensing, Geodesy, Geophysics and Geodynamics. Besides the ‘traditional’ approach of using laser scanning solely as a supplement for photogrammetry in acquiring digital terrain models, ALA has also been applied to various geoscience research problems. After a short overview of airborne laser altimetry activities in Europe, an application of airborne laser scanning dedicated to Alpine Glaciology is presented.
Aircraft laser altimetry measurement of elevation changes of the greenland ice sheet: technique and accuracy assessment
icesheet_lidarassessment.pdf
Krabill, W.B., W. Abdalati, E.B. Frederick, S.S. Manizade, C.F. Martin, J.G., R.N. Swift, R.H. Thomas, Sonntag, R.N., and J.G. Yungel
Journal of Geodynamics 34 (2002) 357–376
Abstract: Airborne laser altimetry has been used during the past decade to measure the surface elevation of the Greenland ice sheet. These measurements have been made using a scanning laser on a NASA P-3 aircraft which was positioned by differential GPS and flown approximately 500 m above the surface. Flights have been made over major portions of the ice sheet and reflown 5 years later in order to obtain estimates of the rate of overall change of surface elevation. The accuracy with which differential elevations can be made depends upon (a) the GPS positioning accuracy, (b) the instrument calibration accuracy, (c) the stability of the laser and, (d) the accuracy of the aircraft inertial navigation system’s estimation of aircraft attitude. Overall, the accuracy of an elevation change estimate is computed to be 8.5 cm over small areas and 7.1 cm when averaged over tens of kilometers as is needed for estimating ice volume changes. This effort supports ±1.4 cm/year resolution for long period surface elevation changes from data acquired which are separated by 5 years. Results of inflight data analyses are consistent with these accuracy estimates.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 45 of 47
9. Intensity Title Extracting Artificial Surface Objects from Airborne Laser Scanner Data
File Name Asconas97.pdf
Author(s) Hug, Christoph
Publication and Annotated Abstract Automatic Extraction of Man-Made Objects from Aerial and Space Images. 2nd workshop, Ascona, Switzerland. 5-9 May 1997
Abstract: An approach to filtering non bare-earth objects using 3D geometry and surface reflectance data is presented. The procedure detects surface objects using a morphological filter. These objects are separated into either buildings or vegetation classes using surface reflectance data, elevation ‘texture’, and surface orientation. Error analysis indicates that surface objects are best detected using reflectance data. A combination of reflectance and surface orientation improves data quality. Detecting and identifying topographic objects in imaging laser altimetry data Investigations Of Airborne Laser Scanning Signal Intensity On Glacial Surfaces - Utilizing Comprehensive Laser Geometry Modeling And Orthophoto Surface Modeling (A Case Study: Svartisheibreen, Norway)
What Affects the Intensity Returns?
Lutz_ALSDD2003.pdf
Hug, Ch., Wehr, A.
IAPRS, 32, 19-26, part 3-2w3.
Lutz, E., Geist, Th., and J. Stotter
Proceedings of the ISPRS working group III/3 workshop ‘3-D reconstruction from airborne laser scanner and InSAR data’. October 8-10, 2003, Dresden, Germany
Abstract: An approach to determine how laser intensity information is affected by, 1) geometry, and 2) surface type, is presented. We show results for the Svartisheibreen glacier in Norway, which indicate that the primary cause of peripheral decreases in intensity values (i.e. cross-path fading) is geometry-dependent. This can be eliminated by developing Composite Laser Intensity Models (CLIMs). Signal intensity is also influenced by the range, surface elevation, and the surface land-cover type. IntensityVariation_ScanHoriz.pdf
O’Hagan, Brett
Scanning the Horizons Newsletter. April/May, 2003. AAM GeoScan. Available at: www.aamgeoscan.com.au
Abstract: An article on Optech’s website (www.optech.on.ca) describes factors which affect ALS measurements. The article is summarized here in a table. Airborne Laser Scanning: Developments in Intensity and Beam Divergence
11arspc_jonas.pdf
Jonas, David
AAM GeoScan, Australia. 2002. Available from: http://www.aamsurveys.com.au/tech_list.htm
Abstract:
While the chief interest in ALS is focused on elevation data, attention is also being directed toward other information collected with ALS systems. This paper explores recent advances with two such features: (1) laser intensity, and (2) beam divergence.
LiDAR Bibliography:
Last updated: 4/29/2004
Page 46 of 47
Title Assessing the Possibility of LandCover Classification Using LiDAR Intensity Data
LiDAR Bibliography:
File Name Landclassification_Graz02
Author(s) Song, J-H., Han, SH, Yu, K., and Y-Il. Kim
Publication and Annotated Abstract Newsletter EuroSDR, 2003 No. 2, S. 11 - 12.
Abstract: We discuss opportunities for applying intensity data for land cover classification. Raw intensity data are rasterized and categorized into different land cover types, including roads, grassland, buildings, and forested land. For this study, however, data acquisition and filtering errors rendered the data too noisy for useful analysis. We describe resampling and filtering methods that can effectively remove noise while preserving most of the original information.
Last updated: 4/29/2004
Page 47 of 47
