论文部分内容阅读
本文提出了一种基于分类矢量量化的极低码率视频编码算法及其VLSI结构,它结合了四叉树分割,分类矢量量化算法,变块大小运动估计等多种算法.在获得较高压缩率同时,又克服了低码率图像编码算法中常见的边缘细节模糊的缺陷.同时,本算法在保留了矢量量化解码器简单的优点的同时,并给出了一种易于VLSI实现的编码器结构.其中运动估计电路可和矢量量化器共用处理器阵列,从而减少了编码器的电路规模.
This paper presents a very low bit rate video coding algorithm based on classification vector quantization and its VLSI structure, which combines quadtree segmentation, classification vector quantization algorithm, variable block size motion estimation and other algorithms.In obtaining a higher compression While at the same time, it also overcomes the defect that the edge detail is common in the low bit rate image coding algorithm.At the same time, this algorithm not only retains the simple advantage of the vector quantization decoder, but also gives an easy VLSI encoder Structure in which the motion estimation circuit shares the processor array with the vector quantizer, reducing the encoder’s circuit size.