Proposed Texture Compression Algorithm Based on Fractal Wavelet Theory

The proposed texture compression algorithm is based on a block-wise approach, where each block is subjected to the modified fractal compression method and is partly represented by Haar wavelet or Laplacian pyramid. In our new method, the advantage of the local fractal compression algorithm is taken for 4*4 blocks compressed independently. To overcome the problem of quad tree structure, we use regular partitioning of compressed block to 2*2 range blocks R. The local domain pool for compressed block consists of only one domain block being the whole compressed block D = B. The resulting fractal code is subjected to further compression using Haar wavelet or Laplacian Pyramid.