中文摘要：

绝大多数视频图像在传输过程中皆要进行混合压缩编码，这种以去掉大量空间性和时间性冗余信息为手段的编码方式将导致解码发生一系列错误，急剧影响恢复后的图像质量．针对这种情况，本文提出了一种改进的基于空域错误隐藏的图像恢复算法，该算法利用已正确接收的邻近宏块的像素值，对当前宏块边缘进行综合分析，运用方向插值算法和五点插值算法计算出宏块内边缘和平坦区域的像素值，从而对直接解码恢复出的像素值进行矫正．从实验结果来看，该算法恢复出的图像质量的PSNR值和计算复杂度皆优于传统双线性插值法和五点插值法．

英文摘要：

The hybrid compression coding method is employed in the transmission of most video images. While the coding method for deleting massive spatial and temporal redundancy information will cause a series of errors in the decoding and lead to a sharp decline of the restored image quality. In such case, this paper puts forward an improved spatial-domain error concealment algorithm. The proposed algo- rithm uses correctly received pixel values of adjacent macro blocks to analyze comprehensively the cur- rent macro-block edge, and applies the directional interpolation algorithm and five point interpolation al- gorithm to calculate inner edge and flat region pixel values of the macro block, and correct the pixel value of the restored ones directly decoded. According to the experiment, in terms of image quality, the value of PSNR and calculation complexity restored through the proposed algorithm are better than that through the traditional bilinear interpolation method and the five point interpolation method.