Fast Fractal Image Coding Using Ambient-Cross Sum of Normalized Block

Fractal image coding has a fatal drawback of being time consuming in encoding process. In response to this problem, this paper proposed an effective method to limit the searching space, which is mainly based on a newly defined concept of ambient-cross sum of normalized block and a related inequality. In detail, after the codebook blocks are sorted according to their ambient-cross sum features, for an input range block being encoded, the encoder uses the bisection search method to find out the initial-matched block (i.e., the domain block having the closest ambient cross sum features to the input range block being encoded). And then the encoder confines efficiently the searching scope of similarity matching to the vicinity of the initial-matched block. Simulation results show that the proposed scheme gives significant improvement in speed and quality as compared to the baseline algorithm with full search.