Progressive Image Compression Using Contourlet Transform

This paper proposes a novel progressive image compression algorithm using the nonlinear contourlet transform. Contourlets are effective in capturing directional information in images using a flexible set of basis functions that are elongated and directional. Since the contourlet transform is redundant, a wavelet based contourlet transform has been applied on the image. The wavelet based contourlet transform is the wavelet transform followed by the contourlet transform on the subbands. The resulting set of coefficients is sparse and has been effectively coded using three different algorithms – Set Partitioning In Hierarchical Trees, Tag Setting In Hierarchical Trees and Bit Plane Coding. The results have been compared based on the peak signal-to-noise ratio for the zoneplate image and it is found to be superior to ordinary wavelet transform