An improved fractal image compression scheme embedding DCT encoder

The fractal technique is efficient in reducing redundancy for encoding images at high compression ratios. However, it is computationally expensive, especially to those images containing many rugged boundaries and fine textures. That is due to the increasing range-to-domain matching operations for small range blocks. A novel method combining fractal and DCT (discrete cosine transform) encoders is presented to improve the performance of the fractal method in both enhancing the quality of the reconstructed images and reducing the computational complexity. This primarily makes use of the capabilities of fractal and DCT encoders for excelling at compressing smooth and complex regions respectively. Extensive simulation results suggest that the PSNR performance of proposed scheme is superior to those of fractal and DCT encoders.