基于非局部均值滤波的视频图像去雨方法  

Method of rain removal using non-local means filter in video image

作　　者：顾巧玲[1] 吴健学[1] 

机构地区：[1]武汉邮电科学研究院,武汉430070

出　　处：《信息技术》2017年第1期131-134,138,共5页Information Technology

摘　　要：提出了一种有效的基于非局部均值滤波的去雨方法。首先利用雨的空间特征检测出视频帧中的雨线区域,并对其进行标记。然后针对当前帧的被标记的像素块使用非雨块匹配算法,用于相邻的帧之间以找出和有雨像素区域相似的块。最后使用时空非局部均值滤波对块中的雨区进行重构,达到去除雨滴的目的。实验结果表明,比起常规算法,文中提出的算法可以更有效地去除雨线,恢复出更高质量的图像。An effective method of rain removal based on non-local means filter is presented. The algorithm firstly detects the possible rain streaks in the video frame,using the spatial features of rain,and mark the detected areas. Then a non-rain block matching algorithm is used for each of the labeled pixels in the current frame. The algorithm is used for the adjacent frames to find the similar blocks with the rain pixel regions. Finally,the analogical blocks are reconstructed by using a temporal non-local mean filter to remove the rain. The experiment results show that compared with the conventional methods,the proposed algorithm can be more effective to remove rain,and restore the quality of the image.

关 键 词：计算机视觉 雨滴检测 雨滴去除 块匹配 非局部均值滤波 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]