Analytical Studies of Multi-Levels Framelet Transform for Image Compression

Abstract

Many recent studies showed that the wavelet transform step consumes a long computing time during image compression process. Also, to obtain high CR, multiple transform levels must be applied. This paper examines the properties and design of multilevels framelet for image compression applications. This work is focused on high CR obtained, with minimally perceptible loss in image quality, at the multi-levels of transforms. The programming language MATLAB is used for implementing the proposed algorithm. Simulated results show that there are a suitable number of levels of transforms based on the nature of image. Higher subbands of framelet decomposition can be eliminated for saving in computing and communication energies. At higher levels of transform, these subbands are being of large effects to improve the image quality, if there are used, at higher CR.