A Fast Fractal Image Compression Method on GPUs

• Main Authors: Jin-Yuan Kuo, 郭晉源
• Other Authors: Chin-Feng Lin
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/07811561590227987477

碩士 === 長榮大學 === 資訊管理學系碩士班 === 100 === Fractal image compression is a powerful and competitive technology that it has not only a high compression ratio but also a good image quality of the reconstructed image. However, fractal image compression rarely used due to its time consuming for computing. In this thesis, we proposed a fast fractal image compression method in a GPU using the CUDA architecture. The mean and variance classification method are used to reduce the number of domain blocks for each range block. The range blocks are evenly assigned to the threads of GPU and generate the minimal RMS values. Our experimental results show that our method has about 30000 times faster than the traditional fractal image compression method and the fractal constructing image of our method also has a good image quality.