基于区域重要性的视频时空分级编码

提出了基于区域重要性的视频时空分级编码方法，即根据人眼视觉敏感程度的不同，将视频中的每帧图像划分为重要程度不同的区域，对每个区域分配不同的时空分辨率，以期 在极低码率的传送环境中能最大限度地保证人眼兴趣区域的图像质量和传送效率。实验结果表明，该方法能够在带限的情况下减少图像的抖动，获得较好的图像主观质量。

Temporal-Spatial Scalable Coding Based on Area Significance

This paper proposed a temporal spatial scalable encoding scheme based on area significance. According to the principle that different parts in video sequence have different perceptual significance to observers, selectively to encode those areas in terms of spatial and temporal resolution to provide good image quality and transmission efficiency of the region of interest in the bandwidth restricted conditions. The coding results prove that this method smoothes the dithering and provides better subject quality compared with the original one.