MPEG-2信噪比可分级图象编码的研究  

The Study on SNR-based Hierarchical Coding of Image in MPEG-2

作　　者：李莉莎[1] 尹昊晖[1] 张家谋[1] 

机构地区：[1]北京邮电大学电信工程学院,北京100876

出　　处：《北京邮电大学学报》1999年第4期50-55,共6页Journal of Beijing University of Posts and Telecommunications

基　　金：国家自然科学基金资助项目!(69672037)

摘　　要：考虑到与标准的兼容及实用推广的前景， 介绍了基于ＭＰＥＧ－２ 国际标准的可分级编码方法的研究．并从ＭＰＥＧ－２ 可分级方法的四种方案中选择了信噪比可分级方法作为主要的仿真方法．信噪比分级方法是一个视频源产生两个具有相同空间、时间分辨率和不同视频质量的视频层，低层自己编码来提供基本视频质量， 增强层则被编码来增强低层．实验结果证明， 该方法的增强层对基层的图象质量确实有相当的增强作用．In view of the compatibility with the standard and the prospect for application's extending, this paper introduced ways of hierarchical coding based on the International standard MPEG 2. We selected the SNR based hierarchical way as the main emulation mode from the four hierarchical ways which are mentioned by MPEG 2. The SNR based hierarchical way produces two video layers from one source video. These two layers have different image quality but the same resolving power in space and time. When decoding , the low layer will be decoded by itself to provide a video with basal quality while the enhanced layer will be used to improve the quality of the video that produced by the low layer. Through experiment, the enhanced effect on the quality of the base layer video is showed to be evident and promising.

关 键 词：MPEG-2 图像编码 信噪比 

分 类 号：TN919.8[电子电信—通信与信息系统]