一种基于预测的矢量量化图像编码方法(英文)  

A Prediction-based Vector Quantization Method for Image Coding

作　　者：石敏[1] 谢胜利[1] 

机构地区：[1]华南理工大学电子与信息学院,广东广州510640

出　　处：《华南理工大学学报（自然科学版）》2006年第1期18-23,共6页Journal of South China University of Technology(Natural Science Edition)

基　　金：国家自然科学杰出青年基金资助项目(60325310);广东省自然科学基金团队项目(04205783);国家自然科学基金资助项目(60505005);科技部重大基础前期研究专项基金项目(2005CCA04100)~~

摘　　要：为了提高图像编码效率,提出了一种基于预测的矢量量化图像编码方法.在编解码端设置维数和大小各不相同的两种码书.对于平滑区域,用较小的码书对输入矢量进行编码,从而加快图像编码速度和降低比特率;对于边缘区域,用较大的码书去量化当前输入矢量以保证图像质量.文中还提出了一种有效的码书设计方法以提高生成的码书的质量.仿真结果表明,相对于其它方法,文中所提出的矢量量化图像编码方法运算速度更快,在同样图像质量下降低了传输过程中所需的比特率.In order to improve the efficiency of image coding, a prediction-based vector quantization method is proposed. Two codebooks with different dimensionalities and sizes are respectively employed at the encoding and the decoding ends. For smooth areas, the input vectors are encoded with small-size codebook to speed up the encoding and reduce the bit-rate. For edge areas, the input vectors are quantized with large-size codebook to maintain the image quality. An efficient method for codebook design is also presented to improve the quality of the resulted codebook. Simulated results indicate that, compared with other decoding methods, the proposed method speeds up the encoding and reduces the bit-rate with the image quality keeping unchanged.

关 键 词：矢量量化 图像压缩 图像编码 码书设计 

分 类 号：TN919.8[电子电信—通信与信息系统]