基于率失真优化图像编码的研究  

Research on image coding based on rate distortion optimization

作　　者：汪涛[1] 

机构地区：[1]哈尔滨商业大学网络与教育技术中心,哈尔滨150028

出　　处：《智能计算机与应用》2017年第3期89-92,共4页Intelligent Computer and Applications

摘　　要：数字图像处理技术的一个非常重要的环节是数字图像压缩技术。其在图像保存和图像传输领域拥有无可替代的作用。矢量量化方法是图像压缩技术中一个重要的组成部分。通过有效地转换图像中像素点,从而高效地对图像进行压缩,为图像的存储和传输提供了可靠的保证。本文详细研究了率失真模型在图像压缩领域的应用。通过高效地结合量化失真和编码码率,从而得到在码率一定的条件,相对更高质量的图像。通过对码率分配策略的研究,根据码率和失真二者的关系推导出矢量量化小波图像编码器的率失真模型,同时将该模型与拉格朗日极值算法有机地结合在一起,从而得到图像不同子带的最优码率分配。本文实验结果显示,根据图像的不同子带的特性,应用本文所提出的的图像率失真模型,可以得到较高质量的还原图像。因此可以得出,通过应用本文的率失真模型,在还原图像的还原质量不下降的条件下,得到更低的图像压缩码率。A very important aspect of digital image processing technology is digital image compression technology. Vector quantization is an important part of image compression. By effectively converting the pixels in the image,the image is compressed efficiently,providing a reliable guarantee for the storage and transmission of the image. In this paper,the application of rate distortion model in image compression is studied in detail. Through the efficient combination of quantization distortion and coding rate,in a certain rate of bit rate,relatively higher quality images are gotten. Based on the study of the code rate allocation strategy,the rate distortion model of the vector quantization wavelet image encoder is deduced according to the relationship between the bit rate and the distortion. At the same time,the model is combined with the Lagrangian extreme algorithm. The optimal rate allocation of different subbands is obtained.The experimental results show that according to the characteristics of different subbands of the image,the image rate distortion model proposed in this paper can obtain the higher quality image. Therefore,it can be concluded that by applying the rate distortion model of this paper,a lower image compression bit rate is obtained under the condition that the reduction quality of the restored image does not decrease.

关 键 词：矢量量化 图像压缩 小波变换 

分 类 号：TN911.73[电子电信—通信与信息系统]