基于动态SMVQ的图像压缩算法  

Image compression scheme based on dynamic SMVQ

作　　者：史红刚[1] 周利莉[1] 陈健[1] 

机构地区：[1]信息工程学院信息科学系,河南郑州450002

出　　处：《计算机工程与设计》2006年第14期2594-2596,2600,共4页Computer Engineering and Design

基　　金：河南省教育厅基金项目(sp200303099)

摘　　要：SMVQ利用相邻图像块之间的相关性来提高编码图像质量。然而,传统的SMVQ当图像边缘位于当前编码图像块右下角时会出现编码误差,为此,研究并提出了一种改进的SMVQ。新算法利用图像所有图像块之间的相关性对当前编码图像块进行预测。对每一个输入图像块,利用边匹配方法从已编码数据中寻找最近块,并用其对输入图像块进行预测,或产生动态码书。然后利用最近块、动态码书或总码书对输入数据进行编码,扩展了编码搜索范围,提高了图像质量。实验结果表明新算法大大降低了比特速率,而且编码图像有较好的视觉效果。The side-match finite-state vector quantizer （SMVQ） exploits the correlations between neighboring blocks （vectors） to improve the quality of the coding image. However, when sides of the image lie in the right-bottom of the current coded image, there is an error. In order to avoid the problem, a new dynamic SMVQ technique is proposed. The new DSMVQ exploits the global interblock correlation of image blocks instead of local correlation in conventional dynamic finite-state vector quantizers （DFSVQs）. For an input block, the closest block is searched from the previously encoded data using side-match technique. The closest block is used as the predication of the input block, or used to generate a dynamic codebook. The input block is encoded by the closest block, dynamic codebook or super-codebook, so the code-vector space is expanded. Experimental results reveal that the new DSMVQ reduces bit rate significantly and provides better visual quality, as compared to the basic VQ and other DFSVQs.

关 键 词：动态边匹配有限状态矢量量化 码书 变比特率编码 边匹配误差 图像压缩 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]