论文信息 - Custom Design of JPEG Quantisation Tables for Compressing Iris Polar Images to Improve Recognition Accuracy

Custom Design of JPEG Quantisation Tables for Compressing Iris Polar Images to Improve Recognition Accuracy

Custom JPEG quantisation matrices are proposed to be used in the context of compressing iris polar images within iris recognition. These matrices are obtained by employing a Genetic algorithm for the corresponding optimisation. Superior matching results in iris recognition in terms of average Hamming distance and improved ROC are found as compared to the use of the default JPEG quantisation table.

[1] John Daugman,et al. Effect of Severe Image Compression on Iris Recognition Performance , 2008, IEEE Transactions on Information Forensics and Security.

[2] Ping-Sing Tsai,et al. JPEG: Still Image Compression Standard , 2005 .

[3] Mo Chen,et al. Modification of standard image compression methods for correlation-based pattern recognition , 2004 .

[4] Gregory K. Wallace,et al. The JPEG Still Image Compression Standard , 1991 .

[5] Andreas Uhl,et al. Comparison of compression algorithms' impact on iris recognition accuracy II: revisiting JPEG , 2008, Electronic Imaging.

[6] Randy P. Broussard,et al. Effects of image compression on iris recognition system performance , 2008, J. Electronic Imaging.

[7] Donald M. Monro,et al. Effects of Sampling and Compression on Human IRIS Verification , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.

[8] Andreas Uhl,et al. Comparison of Compression Algorithms' Impact on Iris Recognition Accuracy , 2007, ICB.

[9] Hyun-Sik Ahn,et al. JPEG Quantization Table Design for Face Images and Its Application to Face Recognition , 2006, IEICE Trans. Fundam. Electron. Commun. Comput. Sci..

[10] Patrick J. Flynn,et al. Comments on the CASIA version 1.0 Iris Data Set , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11] Andreas Uhl,et al. Custom JPEG Quantization for Improved Iris Recognition Accuracy , 2009, SEC.