Research on Lossless Compression Algorithms of Low Resolution Palmprint Images

In this study, lossless grayscale image compression methods are compared on public palmprint image databases. Effect of lossy compression algorithms on biometric samples has been well studied. However, lossless compression algorithms on the compression ratios have little been appreciated. In this study, we review and the stateofart lossless compression algorithms and investigate the performance using different when processing palmprint sample data. In particular, including those based on transformation (integer transform based in the JPEG , JPEG2000 and JPEG XR system, as well as the SPbased transform coding method), based on predictive lossless compression algorithms (LJPEG, CALIC and JPEGLS), dictionarybased compression methods (PNG UHA,7z and RAR). To gain a better and reliable result, these lossless compression algorithms are employed to test on different palmprint databases. Based on the testing results using an open palmprint image database, analysis and comparison, CALIC gives high compression ratios in a reasonable time, whereas JPEGLS is nearly as effective and very fast. The performance shows that a guide is given to choose which lossless palmprint image compression algorithm. At last, to find better solutions on how to improve lossless compression performance, we give some examples and suggestions.