Lossless fractal image compression mechanism by applying exact self-similarities at same scale

Abstract

Almost all of the fractal image compression methods are lossy because the concept behind these traditional fractal compression methods is based on loose self-similarity, i.e., the transformation function applied on a domain block does not lead to exact matching with range block, instead the principle works on the minimum distance between transformed domain block and range block. Unlike the traditional fractal image compression methods the proposed mechanism does not partition the image into domain blocks and range blocks. Only a single partition of image serves the purpose. The proposed mechanism uses the concept that in most of the cases a small portion of an image is replicated in some other places of the same image; therefore need not to be stored more than once. The paper presents a novel algorithm to achieve lossless fractal image compression with less computation and fast compression and decompression process. © 2011 Springer-Verlag.