A new fingerprint image compression based on wave atoms transform

During the last decade the emergence of many transforms called geometric wavelets have attracted much attention of researchers working on image analysis. These new transforms propose a new representation richer than the classical wavelets multiscale representation. This paper presents a comparative study of these transforms in order to determine what is the best transform dedicated to a particular type of image that is the fingerprint image. While we are not trying to compete with the current standard JPEG2000, which is dedicated to all types of images, we are aware that for a particular type of images, we can do better by choosing for this type of specific image, a more appropriate tool than classical wavelets. The results show that for fingerprint images the wave atom offers better performance than the current transform based compression standard.