Fractal Encoding Acceleration And Image Restoration Method Based On The Drawing

This paper presents several new methods to solve problems in fractal image coding and image inpainting which are widely investigated in the field of image processing.Research done in fractal image coding is as follows:1. we first give out the guide lines of the performance for classification methods, and propose a technique called "Adaptive Classification based on the Number of Domains", combined with Mass Center and Satisfied Match, we obtain a fast method for fractal coding. Experiment results indicate that in contrast with exhaustive search, the proposed method can improve the speed of fractal coding much more on foundation of worse quality of decoded image, when the uniformity threshold is 120, encoding time decreased form 131.84 to 3.56; in contrast with original uniform classification, the proposed method improve the speed of fractal coding with just some improvement of quality of decoded image at the same compression ratio, when the uniformity threshold is 20, encoding time decreased about 0.1s,but the quality of the decoded image increased 0.27db;2. In order to improve compression ratio. We adopt quadtree method in fractal encoding, and proposed Different Threshold method. Using Adaptive Classification and Satisfied Match we obtain a faster method for fractal coding based on quadtree; Experiment results indicate that the proposed method improve the speed of fractal coding and obtain high compression ratio with just a little loss of decoded image quality, when the uniformity threshold is 110, encoding time decreased form 122.75 to 2.478, the quality of decoded image just decreased 1.48db; in contrast with original uniform classification, the compression ratio increased 0.56 times.3. Considering square partition ignored the uniformity of image and the large computational complexity of HV partition. On foundation of full analysis of original uniform classification and adaptive classification, we propose a hybrid classification method and use it on Mass Center of image. Then we obtain a fast method for fractal coding based on HV Partition. Experimental results indicate that in contrast with exhaustive search hybrid classification method can improve...