Software framework for testing JPEG quantization tables for iris recognition
Ana Tuba, Jovo Arezina, Raka Jovanović · International Conference on Computer Engineering and Applications · 2012
JPEG is most widely used image compression algorithm. The source of its superb compression ability are quantization tables which determine which frequency components of the DCT transform will be neglected. JPEG standard includes recommended quantization tables for general use. However, specialized applications may need different quantization tables. One such example is biometric iris recognition system where it does not matter whethere the picture of iris looks good but can it be efficiently discriminated from other iris images. This paper describes software framework for testing JPEG quantization tables applied to iris recognition from digital images. It allows for easy testing and inclusion of different comparison modules.