Fractal Image Compression Using Parallel DSP
碩士 === 國立雲林科技大學 === 電機工程系碩士班 === 89 === Fractal image compression is a non-conventional technology developed in the last couple decade. It uses self-similar transform to extract the necessary parameters for image reg- ions; these parameters can then be used to decompress these regions after several...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2001
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/21094873472588539152 |
| Summary: | 碩士 === 國立雲林科技大學 === 電機工程系碩士班 === 89 === Fractal image compression is a non-conventional technology developed in the last couple decade. It uses self-similar transform to extract the necessary parameters for image reg- ions; these parameters can then be used to decompress these regions after several iterations of replacements. The fractal image compression can achieve very high compression ratio, but at a cost of high computation cost for the compression phase. In this thesis, we use high-performance parallel digital signal processors (DSP) to build a parallel processing architecture for our fractal image compression system. The presented fractal image compression system explores various factors to achieve high compression ration and high compression speed. The original image is divided into several subimages to work with our parallel processing architecture; the degradation caused by the division process can be compensated using codebook. The performance of our fractal image compression system is compared with currently existing methods, and the result is promising.
|
|---|