一种新的去除红外图像疵点的加权平均算法  

New Algorithm of Weighted Arithmetic Average for Dead Pixels Correction of Infrared Images

作　　者：张楠[1] 周建勇[1] 唐遵烈[2] 

机构地区：[1]重庆邮电大学通信学院,重庆400065 [2]重庆光电技术研究所,重庆400060

出　　处：《半导体光电》2006年第2期203-205,共3页Semiconductor Optoelectronics

摘　　要：疵点的出现严重地影响了红外摄像器件的成像质量。常用的平均值处理算法或者滤波算法,不但不能有效地处理较大的疵点,反而造成图像失真。提出了一种新的加权平均算法,综合运用标号法和平滑法,并在求均值的过程中对数据加权,处理效果较好,能够有效去除红外摄像系统中的大小疵点,而不影响图像的失真度。The imaging quality of infrared imaging devices is severely effected by the defects. The typical methods such as mean processing algorithm and filtering algorithm can＇t effectively process the larger defects of infrared images, whereas bring about image distortion. A new weighted average algorithm is proposed, in which the data is weighed in the process that the labeling method and smoothing processing method are synthesized. By using this method, the effect of the defects on the infrared image in the infrared imaging system can be effectively eliminated without image distortion.

关 键 词：红外图像 疵点 失真度 标号法 平滑法 加权系数 

分 类 号：TN216[电子电信—物理电子学]