摘要
提出了一种基于小波变换的灰度图像数据压缩方法．采用双正交小波对图像进行分解。作分解后系数的矢量量化。使用频率敏感自组织特征映射算法生成码书．这可避免矢量量化时的块效应，且相对于正交小波，恢复图像的质量也有所提高．
Based on wavelet transform and vector quantization. a scheme for image data compression is presented. Firstly. the image is decomposed by the biorthogonal wavelet filters, and then the wavelet coefficient is encoded by vector quantization. The Frequency Sensitive Self-Organizing Feature Maps (FSOFM) algorithm is used in constructing codebook. A number of computer simulation experimental results show that the blocking' effect can be overcome, and the better reconstructed image can be obtained by the biorthogonal wavelet in comparision with the orthongonal wavelet.
出处
《重庆邮电学院学报（自然科学版）》
2001年第3期46-48,共3页
Journal of Chongqing University of Posts and Telecommunications(Natural Sciences Edition)
基金
重邮青年教师科技基金资助项目
关键词
小波变换
矢量量化
图像压缩
wavelet transform
vector quantization
image compression
