FRACTAL IMAGE COMPRESSION BASED ON DELAUNAY TRIANGULATION AND VECTOR QUANTIZATION

This paper presents a new scheme for fractal image compression based o n adaptive Delaunay triangulation, Such a partition is computed on an initial set of points obtained with a split and merge algorithm in a g rey level dependent way, The triangulation is thus fully flexible and returns a limited number of blocks allowing good compression ratios, M oreover, a second original approach is the integration of a classifica tion step based on a modified version of the Lloyd algorithm (vector q uantization) in order to reduce the encoding complexity, The vector qu antization algorithm is implemented on pixel histograms directly gener ated from the triangulation. The aim is to reduce the number of compar isons between the two sets of blocks involved in fractal image compres sion by keeping only the best representative triangles in the domain b locks set, Quality coding results are achieved at rates between 0.25-0 .5 b/pixel depending on the nature of the original image and on the nu mber of triangles retained.