International Journal of Electronics, Communication & Instrumentation Engineering Research and Development (IJECIERD) ISSN 2249-684X Vol. 3, Issue 4, Oct 2013, 81-88 © TJPRC Pvt. Ltd.

IRIS IMAGE SPECULAR QUALITY ARTIFACT EVALUATION FOR BIOMETRIC RECOGNITION KARAD SACHIN1, SANTOSH GHODAKE2, BHANUDAS YADAV3 & SHAKTIKUMAR SHILEDAR4 1

Assistant Professor, Instrumentation Engineering, Maharshi Parshuram College of Engineering, Velneshawar, Maharashtra, India 2

Senior Lecturer,Electronics Engineering Department,Government Polytechnic,Mumbai, Maharashtra, India 3

Head of Department,Medical Electronics Government Polytechnic Nanded, India

4

Incharge, Head of Department,Electrical Engineering,Government Polytechnic,Mumbai, Maharashtra,India

ABSTRACT Image quality assessment plays an important role in the performance of biometric systems. Data quality assessment is a key issue, in order to broaden the applicability of iris biometrics to unconstrained imaging conditions. Having empirically observed the published strategies to assess iris image quality; In this paper, we propose quality factor after assessing the prominent factors of iris like Dilation, Light variation, Occlusion and Specular reflection by their scores. Mainly Concentrating on Specular quality artifact. These factors once evaluated are then used to get quality score for a given iris image from a database. As far as possible we have ensured that this paper will follow a common protocol in deciding those features. Not only in finding the feature from one database but can be implemented and quality score can be assessed for all possible freely available databases. Initial work has been carried out from the database created by our own set up (HUWITZ HS 5000).The comparison is also done with existing databases and algorithm. This in turn will act as a benchmark in increasing the efficiency of further processing.

KEYWORDS: Biometrics, Image Quality Assessment, Iris Image Quality, Iris Recognition, Quality Metrics INTRODUCTION Security and the authentication of individuals is necessary for many different areas of our lives, with most people having to authenticate their identity on a daily basis; examples include ATMs, secure access to buildings, and international travel. Biometric identification provides a valid alternative to traditional authentication mechanisms such as ID cards and passwords, which overcoming many of the shortfalls of these methods; it is possible to identify an individual based on ―who they are" rather than what they possess" or ―what they remember. Iris recognition is a particular type of biometric system that can be used to reliably indentify a person by analyzing the patterns found in the iris. The iris is so reliable as a form of identification because of the uniqueness of its pattern. Although there is a genetic influence, particularly on the iris' colour, the iris develops through folding of the tissue membrane and then degeneration (to create the pupil opening) which results in a random and unique iris. In comparison to other visual recognition techniques, the iris has a great advantage in that there is huge variability of the pattern between individuals, meaning that large databases can be searched without finding any false matches. This means that irises can be used to identify individuals rather than just confirm their given identity; a property that would be useful in a situation such as border control, where it might be important not to just show that an individual is not who they say they are but also to show exactly who they are. The objective of this project is to produce a working prototype program that functions as an iris image quality assessment tool

82

Karad Sachin, Santosh Ghodake, Bhanudas Yadav & Shaktikumar Shiledar

using the algorithms described by Professors John Daugman and Hugo Proenca and also to implement this in an accurate and useful way that is user-friendly. IRIS IMAGE quality assessment plays an important role in automated biometric systems for two reasons: (i) system performance and (ii) interoperability. Due to the effectiveness proven by the deployed iris recognition systems, the popularity of the iris biometric trait has considerably grown in the last few years. Among these biometric methods, iris is currently considered as one of the most reliable biometrics because of its unique textures random variation. Moreover, iris is proved to be well protected from the external environment behind the cornea, relatively easy to acquire and stable all over the person‘s life. For all of these reasons, iris patterns become interesting as an alternative approach to reliable visual recognition of persons. Several reasons justify this interest:1) it is a naturally protected internal organ that is visible from the exterior; 2) it has near circular and planar shape that more easily turns its segmentation and parameterization; and 3) its texture has a predominantly randotypic chaotic appearance that is stable over a lifetime. Biometric methods, which identify people based on physical or behavioral characteristics, are of interest because people cannot forget or lose their physical characteristics in the way that they can lose passwords or identity cards. Previous work on iris image quality can be placed into two categories: global and local analysis. Global image assessment is performed on the entire image and does not use intrinsic Information specifically pertaining to the iris, while local analysis uses only information related to the iris, which requires iris and/or pupil localization. Global analysis is typically faster than local analysis because it does not involve iris localization. However, it is much more difficult, if not impossible, to get an accurate assessment of the iris through global analysis because non iris regions such as the sclera, eyelids, eyelashes, and eyebrows negatively influence the assessment. On the other hand, local assessment can provide a much more accurate analysis of the iris but only with correct localization. To this end, careful consideration must be placed in the localization algorithms chosen for this type of assessment. Iris recognition is a particular type of biometric system that can be used to reliably identify a person by analyzing the patterns found in the iris. The iris is so reliable as a form of identification because of the uniqueness of its pattern. Although there is a genetic influence, particularly on the iris' color, the iris develops through folding of the tissue membrane and then degeneration (to create the pupil opening) which results in a random and unique iris. Both enrollment and matching process include image acquisition, iris localization, iris normalization and feature extraction. In enrollment process, extracted feature vector is stored in the database. During the matching, the extracted feature is compared with stored features. Iris localization is the most important step in iris recognition system. The performance of the iris recognition system depends on the precision of the localization step. The cost (execution time) of localization is more than half of the recognition time, therefore, perfection of the localization as well as the time complexity of the localization step are challenging issues in iris recognition system. Iris localization is one of the crucial parts in an excellent iris recognition system, which should be able to cope with many bad circumstances such as blurred outer boundary of iris, those light points in pupil and the effects caused by eyelids and eyelashes. The boundaries of iris region can be approximated using two circles, one for the iris/sclera boundary and another, interior to the first, for the iris/pupil boundary. The aim of iris location is to estimate the centre as well as radius of the two circles.

PROPOSED SCHEME One of the open problems in biometrics that limits the applicability of iris biometrics is iris identification at-adistance and on-the move that must be done in unconstrained imaging conditions and using potentially large databases. To broaden the usability of iris biometrics, additional analysis techniques must be developed, taking into account the specificity of the degraded images, such as light reflections from the eye‘s surface, occlusions and fluctuations of perspective and illumination. Most of these factors limit availability of distinctive features needed for proper recognition.

Iris Image Specular Quality Artifact Evaluation for Biometric Recognition

83

Thus, recently, significant effort has been focused on authenticating objects at-a-distance and on-the-move using the iris biometrics on the basis of quality. The performance of Iris Recognition is affected by poor quality of iris images. Thus Iris image quality assessment plays a major role which is based on analyzing the effect of quality factors in which primary factors are defocus blur, motion blur, off-angle and secondary factors are Specular reflection, occlusion, lighting, pixel count, Dilation Score, Lightning Variation. In this paper we propose quality factor after assessing the prominent factors of iris like Dilation score, Light variation, Occlusion score and specular reflection score. These factors are evaluated from 15 databases using Proposed Algorithm. Initial work has been carried out from the database created by our own set up (COEP Data Base). The comparison is also done with existing databases and algorithm.

Figure 1: Block Schematics for Lightning

Figure 2: Block Schematics of Specular Reflection

Figure 3: Block Schematics for Occlusion Score

84

Karad Sachin, Santosh Ghodake, Bhanudas Yadav & Shaktikumar Shiledar

Figure 4: Block Schematics for Dilation Score

IMPLEMENTATION Dilation Measure In addition to occlusion, the dilation of a pupil can affect the recognition accuracy. If the iris is too dilated, there would not be enough information for recognition. In this paper, the dilation measure (D) is calculated by (1) Specular Reflection Once eyelid occlusions are estimate occlusions resulting from Specular reflections are evaluated on the remaining iris portion unaffected by the eyelids. (2) Occlusion Measure The total amount of available iris pattern scan affects the recognition accuracy. Daugman used 50% as a threshold. If the occlusion is more than 50%, then the iris image will be considered poor quality and is not used for recognition. This is a hard threshold. In this paper, we developed the occlusion measure. The occlusion measure (O) is to measure how much percentage of the iris area is invalid due to eyelids, eyelashes, and other noise. (3) Lighting Variation After estimating occlusions from eyelids and specular reflection, the remaining unconcluded iris portion is split into5 regions. A mean is calculated in each region (Xi) (4)

RESULTS AND DISCUSSIONS In this paper, we proposed the feature information approach for iris image-quality measure for iris recognition. The quality score can be used to calculate the confidence level of the. The quality score can be used to calculate the confidence level of the recognition result. Three public accessible databases are used in our experiment: 1) CASIA (v1, v2 s1, v2 s2, 3L, 3T, 3I); 2) UB Iris (800600s1, 800600 s2, 20 150, 200 1502, 200 150 R1, 200 150 R2); 3) MMU (1, 2); 4) UPOL; 5) COEP. These databases cover a wide range of iris image types and allowed thorough testing of our method. We analyzed four quality score Lightning Variation, Speular Reflection, Dilation Score and occlusions score component and

85

Iris Image Specular Quality Artifact Evaluation for Biometric Recognition

their distributions using these databases. Our analysis shows that the proposed quality measure is consistent with our observations. Specular Reflection Quality Score Analysis Specular Reflection: Once eyelid occlusions are estimated, occlusions resulting from specular reflection are evaluated on the remaining iris portion unaffected by the eyelids (5) We Calculated result on different 15 databases as mentioned above are as Table 1: Specular Score Analysis Database CAS 3I CAS 3L CAS 3T CAS V1 CAS V2 S1 CAS V2 S2 UB 200 150 UB 200 1502 UB 200 150 R1 UB 200 150 R2 UB 800600 S1 UB 800600 S2 MMU DB 1 MMU DB 2 COEP DB

Minimum 0.0031 0.0016 0.0342 0.0944 0.0044 0.0001 0.1531 0.0081 0.1936 0.0052 0.1663 0.0057 0.0706 0.0009 0.0197

Maximum 0.9989 0.9983 0.9542 0.9960 0.9974 0.9987 0.8979 0.9758 0.9351 0.9041 0.8510 0.8751 0.9750 0.9958 0.3613

Average 0.8753 0.5753 0.5394 0.7985 0.4338 0.3665 0.4571 0.4173 0.4859 0.3565 0.4853 04074 0.5029 0.4454 0.1552

Standard Deviation 0.1235 0.2400 0.2303 0.1396 0.2723 0.2705 0.1247 0.2088 0.1345 0.2048 0.1461 0.1968 0.1849 0.2404 0.0663

Table 2: Specular Score Interval Database CAS 3I CAS 3L CAS 3T CAS V1 CAS V2 S1 CAS V2 S2 UB 200 150 UB 200 1502 UB 200 150 R1 UB 200 150 R2 UB 800600 S1 UB 800600 S2 MMU DB 1 MMU DB 2 COEP DB

Low 0~0.0031 0~0.0016 0~0.0342 0~0.0944 0~0.0044 0~0.0001 0~0.1531 0~0.0081 0~0.1936 0~0.0052 0~0.1663 0~0.0057 0~0.0706 0~0.0009 0~0.0197

Medium 0.9989 0.9983 0.9542 0.9960 0.9974 0.9987 0.8979 0.9758 0.9351 0.9041 0.8510 0.8751 0.9750 0.9958 0.3613

High 0.9989~1 0.9983~1 0.9542~1 0.9960~1 0.9974~1 0.9987~1 0.8979~1 0.9758~1 0.9351~1 0.9041~1 0.8510~1 0.8751~1 0.9750~1 0.9958~1 0.3613~1

86

Karad Sachin, Santosh Ghodake, Bhanudas Yadav & Shaktikumar Shiledar

Figure 5: Histogram of Dilation Score

CONCLUSIONS AND FUTURE SCOPE In this research, we proposed the Quality Estimation approach for iris image-quality measure for iris recognition. The quality score can be used to calculate the confidence level of the recognition result. Sixteen public accessible databases are used in our experiment: 1) CASIA (v1, v2 s1, v2 s2, 3L, 3T, 3I); 2) UB Iris (800600s1, 800600 s2, 20 150, 200 1502, 200 150 R1, 200 150 R2); 3) MMU (1, 2) ; 4) UPOL; 5) COEP . These databases cover a wide range of iris image types and allowed thorough testing of our method. We analyzed four quality score Lightning Variation, Specular Reflection, Dilation Score and occlusions score component and their distributions using these databases. The analysis shows that the proposed quality measure is consistent with our observations. The main limitation of this approach is the requirement of segmentation. Failed localization/segmentation will result in inaccurate quality scores. Therefore, as long as the segmentation algorithm used for quality evaluation is as sophisticated as the one used in quality evaluation, nevertheless, the need to deploy segmentation within the quality assessment algorithm makes this approach unsuitable for real-time applications in which a quality factor would be used for the selection of the “best” frame from a sequence. Future work includes perfecting the estimation techniques for the described quality factors, along with experimenting with the new quality scores that incorporate correlation. Furthermore, the proposed framework is open for the inclusion of new iris quality factors that will undoubtedly emerge through further research or through further relaxation of acquisition constraints (e.g., distance, motion, and non uniform lighting).

REFERENCES 1.

N. Kalka, J. Zuo, N. Schmid and B. Cukic., “Image quality assessment for iris biometric.”, Proceedings of SPIE The International Society for Optical Engineering, v 6202, 2006, Biometric Technology for Human Identification III, may 2010.

2.

A. Abhyankar and S. Schuckers, “Iris quality assessment and bi-orthogonal wavelet base decoding for recognition,” Pattern Recognition, vol. 42, pp. 1878–1894, 2009.

Iris Image Specular Quality Artifact Evaluation for Biometric Recognition

3.

87

Hugo Proença “Quality Assessment of Degraded Iris Images Acquired in the Visible Wavelength”, IEEE Trans. on Information Forensics and Security, Vol. 6, NO. 1, March 2011.

4.

X. Zhu, Y. Liu, X. Ming, and Q. Cui, “A quality evaluation method of iris images sequence based on wavelet coefficients in ’region of interest’,” Proc. of the 4th Int’l Conf. on Computer and Information Technology, pp. 24– 27, September 2004.

5.

Y. Chen, S. Dass, and A. Jain, “Localized iris quality using 2-D wavelets,” International Conference on Biometrics, pp. 373–381, 2006. Hong Kong, China.

6.

L. Ma, T. Tan, Y. Wang, and D. Zhang, “Personal identification based on iris texture Analysis,” IEEE Trans. on Pattern Analysis and Machine Intelligence 25, pp. 1519–1533, December 2003.

7.

B. Kang and K. Park, “A study on iris image restoration,” Proc. Audio and Video-Based Biometric Person Authentication 3546, pp. 31–40, July 2005.

8.

J. Daugman, “How iris recognition works,” IEEE Trans. on Circuits and Systems for Video Technology 14, pp. 21–30, January 2004.

9.

C. Shi and L. Jin., “A fast and efficient multiple step algorithm of iris image quality assessment”, Proceedings of the 2010 2nd International Conference on Future Computer and Communication, ICFCC 2010, v 2, p V2589V2593, 2010, Proceedings of the 2010 2nd International Conference on Future Computer and Communication, ICFCC 2010.

10. R. Wildes, “Iris recognition: An emerging biometric technology”, Proceedings of the IEEE 85, pp. 1348 –1363, September 1997. 11. W. Boles, B. Boashash, “A human identification technique using images of the iris and wavelet transform,”IEEE Transactions on Signal Processing, Vol. 46, No. 4, 1998. 12. Y Ping Huang, “An Efficient Iris Recognition System,” First International Conference on Machine Learning and Cybernetics, Beijing, 4-5 November 2002 13. T. Camus and R. Wildes, “Reliable and Fast Eye Finding in Close-up Images,” 16th International Conference on Pattern Recognition, vol. 1, 2002, pp. 389-394. 14. L. Ma, T. Tan, Y. Wang, and D. Zhang, “Personal Identification Based on Iris Texture Analysis,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 25, no. 12, December 2003, pp.1519-1533.