Fast fractal image encoding based on correlation information feature

This paper presents a fast fractal encoding approach based on correlation information feature. To reduce the computational complexity of fractal encoding, the definition of correlation information feature is introduced and used to convert range-domain block matching to nearest neighbor search in the feature space. Correlation information feature can reflect the blocks´ structural characteristic well. Compared with conventional features mainly reflecting blocks´ statistical characteristic, correlation information feature can provide better searching results. Experiments show that the proposed algorithm yields superior performance over two other fast fractal encoding algorithms.