论文部分内容阅读
提出一种基于小波变换的图像数据的压缩编码方法 ,基本思想是利用小波变换实现图像的多分辨率分解 ,采用了结合矢量量化 (VQ)的混合编码技术对分解后的图像进行编码 ,考虑图像的边缘轮廓对视觉质量的影响 ,利用了小波图像的方向选择性和低频子带的边缘信息组织码本 ,在矢量量化 L BG算法的初始码书的选取中根据矢量中各分量的特性提出一种改进的随机选取法 ,避免了可能的胞腔不均现象 ,提高了码书的质量 ,重构的图像质量也有所提高
A method of compressing and encoding image data based on wavelet transform is proposed. The basic idea is to use wavelet transform to realize multi-resolution decomposition of images. A hybrid coding technique based on vector quantization (VQ) is used to encode the decomposed image, The influence of the edge contour on the visual quality is analyzed. By using the direction selectivity of wavelet image and the edge codebook of low frequency subband, the initial codebook of vector quantization L BG algorithm is proposed according to the characteristics of each component in the vector Improve the random selection method, to avoid the possible phenomenon of uneven cells, improve the quality of the codebook, reconstructed image quality has also been improved