小波-分形混合图像编码算法中双正交小波的选择

THE CHOICE OF THE WAVELET FOR WAVELET- FRACTAL HYBRID IMAGE CODING ALGORITHM

分析了分形变换对图像中纹理区域的编码机理以及小波变换的一些特征参数对编码效果的影响，同时根据小波变换作为ＫＬＴ 变换的近似变换所具有的能量聚集特性，构造了一个新的小波变换特征参数ＭＳＲ（主旁比），进而分析了ＭＳＲ对编码效果的影响，并得出结论：ＭＳＲ是进行针对小波－分形混合编码中较好的双正交小波选择的一种有效判别参数．

This paper analyzes the mechanism of the fractal transform acting in the texture area of natural images and the influence of some features of wavelet transform on coding performance and constructs a new wavelet feature MSR(Main wave to Side wave Ratio) which lies in the fact that the wavelet transform acts as an approximate KLT transform for energy concentration. This paper also suggests that MSR is the crucial feature to choose the good wavelet for wavelet fractal image coding.