Two Fast Encoding Algorithms For Fractal Image Compression

Fractal image coding is a promising and novel image compression technique. In recent ten years, it has received much attention due to its high compression ratio, multi-resolution and very fast decompression. Fractal image coding opens up a refreshing new view to image compression.The excellent visual quality and high compression rate of fractal image coding have the limited applications due to exhaustive inherent encoding time. Therefore, this dissertation is focused on how to reduce runtime in the encoding process while maintaining the reconstructed image quality.This dissertation proposes two fast fractal encoding algorithms. They are summarized as follows:①Algorithm based on a semi-norm of normalized block (Chapter 4). Because the runtime in the encoding process is mostly spent on searching for the best-matched block to an input range block in a usually large domain pool, this fast algorithm is proposed to limit the search space; it can find out the best-matched block to an input range block in a relatively-small search neighborhood. The proposed algorithm has a better performance in terms of encoding time and image quality, compared with the fast algorithm based on the cross-trace [Improving cross trace-based algorithm for fractal image coding, Chinese Journal of Computers, 2007, no.10].②Algorithm combined with wavelet transform (Chapter 5). Unlike the existing feature-type algorithms that use only one feature to describe all the image blocks, the current algorithm is combined with the idea of blocks classification. In detail, it is first to classify image blocks into three categories using wavelet transform, then uses a feature-type algorithm based on each of three newly defined features in each category. The experiments show that the proposed algorithm performs better in terms of encoding time and image quality, compared with the some existing feature-type algorithms.