一种基于运动估计的3D视频降噪算法  

3D Video Noise Reduction Based on Motion Estimation

作　　者：龙红梅[1] 田逢春[1] 谭洪涛[1] 任晓羽[1] 

机构地区：[1]重庆大学通信工程学院,重庆400030

出　　处：《计算机与数字工程》2009年第6期122-124,148,共4页Computer & Digital Engineering

摘　　要：在视频图像降噪中,时域滤波比空域滤波在保护边缘和细节,提高PSNR方面更具有优势。根据此原理,提出了一种基于运动估计的3D降噪算法。该3D降噪算法结合了时域滤波和空域滤波。在时域上,基于运动估计在当前帧的前一帧和后一帧中同时搜索匹配块。对搜索到的匹配块进行运动强度检测,如果运动强度较小,就进行时域滤波,如果运动强度过大,就为对当前块进行空域滤波。另外,设计了噪声标准差估计单元,能够根据噪声标准差自动调整运动强度检测阈值,准确判断块的运动强度。同时,估计出的噪声标准差也用作空域滤波器的参数。Temporal filter has the advantage over spatial filter in the protection of edge and detail, as well as the enhancement of PSNR. A 3D video noise reduction Based on motion estimation is proposed. This 3D filter for video noise reduction combines spatial and temporal filter together. Matching block of current block is searched in the former frame and next frame based on motion estimation. Then, the motion between current block and matching block is detected. If the motion is smaller than the threshold, a temporal filter is used, otherwise, a spatial filter is used rather than temporal filter. Besides, a noise standard deviation detecting unit is used to get the threshold of motion detect and the parameter of spatial filter.

关 键 词：视频图像降噪 运动估计 时域 空域 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]