论文部分内容阅读
提出了一种基于ＬＢＧ码书设计的新的图像矢量量化算法。该算法利用图像信号在正交矢量空间中的能量集中性，有效地减小了码书的搜索范围，加快了矢量量化速度。同时利用原始图像和重建图像之间的差值进行了自适应补偿，在保证较高压缩比的同时，有效地克服了矢量量化的致命缺陷，即重建图像存在严重的方块效应。
A new image vector quantization algorithm based on LBG codebook design is proposed. The algorithm uses the energy concentration of the image signal in the orthogonal vector space, effectively reduces the search range of the codebook and speeds up the vector quantization. At the same time, adaptive compensation is made by using the difference between the original image and the reconstructed image, which can effectively overcome the fatal flaw of vector quantization while ensuring a high compression ratio. That is, the reconstructed image has a serious box effect.