Image compression using multidirectional anisotropic transform: Shearlet transform

It is now widely acknowledged that the geometrical features of the images are essential for many applications including image compression. A novel transform coding technique based on shearlet transform is proposed. A multi directional anisotropic shearlet transform is used to detect the prominent features like edges of the image. Indeed, unlike traditional wavelets, shearlets are theoretically optimal in representing images with edges and other geometrical features. Thresholding method is applied to shearlet transform coefficients and finally threshold output is encoded using Set Partitioning in Hierarchical Trees (SPIHT) technique. Compare to wavelet based compression technique; the proposed shearlet based compression technique is more efficient for wide range of geometrical features of the images. The simulation results also show that this new compression technique is competitive to the wavelet based compression method in terms of the PSNR and visual quality of the reconstructed image.