基于运动补偿的红外图像噪音时域ⅡR滤波算法  

Time Domain ⅡR Filtering Algorithm for Infrared Image Noise Based on Motion Compensation

作　　者：钱惟贤[1] 陈钱[1] 顾国华[1] 管志强[1] 

机构地区：[1]南京理工大学近程高速目标探测国防重点学科实验室,南京210094

出　　处：《光子学报》2009年第6期1512-1515,共4页Acta Photonica Sinica

基　　金：兵器科技预研项目(30404010103)资助

摘　　要：从模式识别的角度出发将图像分为运动区域和非运动区域,采用基于噪音分布模型的运动检测技术与基于概率松弛法的运动区域标号技术相结合对这两种区域进行具体划分,对分类后的图像采用时域ⅡR滤波,不同区域采用不同的影响因子,以期达到最终的运动补偿.仿真结果表明该方法有效抑制了噪音,同时并未引起明显的运动模糊,图像的视觉效果得到了极大改善.To solve the blur of infrared image caused by the phase delay of the time domain filter, pattern analysis theory was used to divide the image into two regions,while one is the motion region,and the other is the non-motion region. And the technique of the motion detection based on noise probability and the technique of the region label based on probability relaxation were used for the region dividing. For different motion regions,the IIR filters with different coefficients were used to achieve motion compensation. The resuh shows that this technology can achieve noise removal and blur protection, and the vision effect is much better.

关 键 词：时域滤波 相位延时 运动模糊 运动检测 概率松弛法 ⅡR滤波 

分 类 号：TN911.73[电子电信—通信与信息系统]