SPIHT Classification Based Fast Fractal Image Coder

Embedded zero-tree coding in wavelet domain has drawn a lot of attention to the image compression applications. Among noteworthy zero-tree algorithms is the SPIHT algorithm. For images with textures, high frequency wavelet coefficients are likely to become significant after few scan passes of algorithms and therefore the coding results are often need to be sufficient. This paper presents a new fast and efficient image coder that applies fractal image encoding method based on classification of image blocks is presented. Two different parameters are used as classes to sort image blocks into: the direction of the approximate first derivative and a normalized root mean square error of the fitting plane for the given block is applied to wavelet transformed image of the low pass sub band and a modified set partitioning in hierarchical trees (SPIHT) coding, on the remaining coefficients. These parameters are the number of domain blocks and examined for a range block is reduced dramatically. This classification results in a considerable acceleration of the encoding process, without loss of the reconstruction fidelity. The proposed method is compared to its rate-distortion performance under the same encoding time limit,and compared with full search practical swarm optimization then compared recently developed fast classification algorithms such as ‘No search algorithm’, are proved to be better and average of 94% reduction in encoding- decoding time comparing to the other Fractal coding results.