一种基于标量量化分割的视频多描述编码方法  

A video multiple description coding approach based on scalar quantization division

作　　者：蔡旻[1] 欧扬[1] 胡伏原[1] 雷岩[1] 

机构地区：[1]苏州科技学院电子与信息工程学院,江苏苏州215009

出　　处：《苏州科技学院学报（自然科学版）》2015年第4期46-50,共5页Journal of Suzhou University of Science and Technology （Natural Science Edition）

基　　金：国家自然科学基金资助项目(61472267);江苏省自然科学基金资助项目(BK2012166)

摘　　要：多描述编码技术能提高编码的鲁棒性来保证视频流的高效稳定传输,是一种重要的差错复原技术。结合多描述编码理论和视频编码具体情况,提出了一种基于标量量化分割的视频多描述编码方法,通过归纳得到两张索引匹配表,并由此将DCT系数分割为两路,实现了在编码端的分割。而在接收端通过增加反映射模块,实现每一路都能单独解码,两路全部收到可以联合解码。最后通过实验证明:该方法是可行的,并取得比较好的图像效果。Multiple Description Coding （MDC） technology belongs to error-resilience, which ensures the good transmitting of video stream by enhancing the robust of coding. Considering MDC theory and situations of video coding, this paper proposes a video multiple description coding approach based on scalar quantization division by concluding two tables of index matching. In this way, we divide DCT coefficients into two pathways at the cod- ing-end. At the decoding-end, each bit-stream can be decoded independently by adding the reverse module, joint decoding when both bit-streams were received. The result of the experiment shows that this approach is fea- sible, and can obtain good quality of the video.

关 键 词：视频编码 多描述编码 多描述标量量化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]