基于视觉敏感的视频差错复原算法  

Error Resilience Scheme Based on Image Segmentation

作　　者：肖明明[1] 梁凡[1] 

机构地区：[1]中山大学电子与通信工程系

出　　处：《计算机工程与应用》2006年第26期173-175,200,共4页Computer Engineering and Applications

基　　金：国家863高技术研究发展计划资助项目(编号:2004AA119010);广东省自然科学基金资助项目(编号:031546)

摘　　要：根据人眼更多的注意感兴趣的区域和倾向对背景的改变不太敏感的视觉特性,提出基于视觉敏感图像分割的视频差错复原算法,实现支持该算法的H.264编解码器。利用灵活宏块排序技术(FMO)对图像帧进行分割,将图像帧分割为前景子图像和背景子图像。与背景子图像相比,对前景子图像按更高图像质量的要求进行编码,并且对前景子图像的传输进行更有效的保护。实验结果表明,提出的基于图像分割的差错复原算法具有良好的容错性能,重建图像的主客观质量良好。In general,people think a lot of on foreground of picture when they look at the picture.So an image segmentation based error resilience algorithm is introduced.The video frame is divided into foreground sub-picture and background sub-picture with flexible macroblock ordering tool（FMO）,then some fine code is done to foreground subpicture,and transmission error control is implemented by forward error control packet to code data of foreground subpicture.Simulation results show that the overall subjective quality is considerably improved compared to the conventional coding schemes.

关 键 词：见频压缩编码 差错复原技术 灵活宏块排序 图像分割 

分 类 号：TN919.81[电子电信—通信与信息系统]