13th SpectroNet Collaboration Forum

Innovative/Smart Color and Spectral Imaging in Industry, Biology and Medicine

Innovative Image Processing Technologies for Quality Assurance in Industry Prof. Dr.-Ing. Gerhard Linß Dr.-Ing. Maik Rosenberger Dipl.-Ing. Marc Preißler Ilmenau University of Technology Faculty for Mechanical Engineering Department of Quality Assurance and Industrial Image Processing 21.04.2012 page 1 www.tu-ilmenau.de/qualimess

1. State of the art in multispectral image processing

Aeriel view of Sylt [University of Kiel}

Fusarium infection [VideometerLab]

21.04.2012 page 2 21.04.2012 www.tu-ilmenau.de/qualimess

wavelength

Seite 2/21

1. State of the art in multispectral image processing Colorimetry

multispectral imaging device [can:scan from caddon]

JENCOLOR Multiple Color Sensor [MAZet Jena GmbH]

luminance measuring videophotometer LMK 98-4 [Techno Team Bildverarbeitung GmbH]

 fast color measurement and spectral analysis  virtual reproduction of a sampled surface  color measurement mostly based on spectral evaluation and selective spectral analysis 21.04.2012 page 3 www.tu-ilmenau.de/qualimess

1. State of the art in multispectral image processing Key Features

Line scanner

Multiple chip cameras

Filter wheel cameras

Liquid crystal tunable filters cameras

Spectral filter range (approx.)

400-1000nm

400-1000nm

Depends on Filter technology

400-2450nm

Single channel bandwidth

ca. 70nm

ca. 100nm (VIS)

ca.5 nm

ca. 0,25nm

Modularity

Static sensor, static filter

Static sensor, static filter

Changeable sensor and filters

Changeable sensor, tunable filters Tunable (almost unlimited)

Number of channels

5-12

5

Depends on diameter, average 56

Pixel displacement error between different channels

-

-

none

none

Sampling mode

Parallel, line scan

Parallel, area scan

Sequential, area scan

Sequential, area scan

Sampling speed

fast

fast

slow

slow

Price

>7000€

>30000€

>10000€

>45000€

21.04.2012 page 4 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 4/21

2. Investigation of a new filterwheel camera design  Spectrum detection by use of up to 12 exchangeable color channels (400nm – 950nm)

 Interchangeability of two optoelectronic Sensors  Adjustment of motion imbalance  CMOUNT Standard  Possibility for embedded image preprocessing

2

 Free access to the dynamic parameters  Support of high precision and high speed mode

1

 High speed direct drive with position feedback for Precision Mode  Maximum of 2500 rpm and 40 multispectral full frames in Dynamic Mode  Adjustment of motion imbalance  Possibility to replace the bandpass interference filters

21.04.2012 page 5 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 5/21

2. Investigation of a new filterwheel camera design Overview of the spectral sensitivities

 sensors and filters are adaptable 21.04.2012 page 6 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 6/21

3. Internal signal processing  option for embedded image preprocessing (Spartan 3 1800 ADSP FPGA ) µC Trigger signals

Image data

processed image data

FPGA

 Unique color channel assignment for the corresponding singlechannel image

21.04.2012 page 7 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 7/21

3. Tasks for internal signal processing  dynamic adaption of color channels for exposure and gain settings (FPGA)

650nm 600nm

 image registration for static pixel offset  image registration of moving objects

21.04.2012 page 8 Seite 21.04.2012 8/21 www.tu-ilmenau.de/qualimess

4. Principle control modes for users

21.04.2012 page 9 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 9/21

4. Control mode for users Visualizing and mapping user interface  mapping of pseudo color and RGB images  white balance factor settings

 live image display and processing of stored images  spectral side bar for fast monitoring of interesting pixels

21.04.2012 page 10 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 10/21

4. Control mode for users Controlling of the image channels  examples of pseudo color and RGB images recorded with the filter wheel camera

700nm channel

700nm + 766nm channel

766nm channel

RGB

21.04.2012 page 11 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 11/21

5. Application areas Lens inspection for the use in human eye surgery

 Investigations on most suitable contrast detection in single spectral color channels

21.04.2012 page 12 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 12/21

5. Application areas Lens inspection for the use in human eye surgery contrast detection ICX 424 26

21

contrast

16

Inner edge

11

Outer edge 6

1 400

500

-4

600

700

800

900

wavelength in nm

 cut off at 650nm  contrast maximum at 766nm filter

21.04.2012 page 13 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 13/21

1000

5. Application areas Analysis of various fiber-reinforced plastic pipes  various fiber-reinforced plastic pipes for enhance stability

21.04.2012 page 14 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 14/21

5. Application areas Analysis of various fiber-reinforced plastic pipes  inspection of integrated fiber layer  high demands in variety of colors

21.04.2012 page 15 21.04.2012 www.tu-ilmenau.de/qualimess

Seite 15/21

Thank you for your attention! supported by

21.04.2012 page 16 www.tu-ilmenau.de/qualimess