摘要
提出了一种基于运动图像分析的红外焦平面阵列疵点补偿算法,它利用序列图像帧间的相关性,通过对相邻两帧图像进行运动分析,获取图像中目标的运动参数,再通过运动参数,将当前图像中的疵点对应到前一帧图像相应的位置,并用前一帧相应位置像素的灰度对疵点进行补偿.该算法克服了邻域疵点补偿算法无法保持目标边缘的缺点,其补偿效果优于邻域补偿算法的效果.
This paper presents a new blind pixels approximation algorithm based on the scene moving analysis for the infrared focal plane array（IRFPA）. It is known that there exists strong relativity between neighboring images in image sequences. By moving scence analysis, moving parameters can be calculated from image sequences. Then the pixels in the previous frame image can be found which correspond to blind pixels in the current frame image. And these pixels are used to approximate blind pixels in the current frame. This algorithm can make objects in the aproximated image maintain its edge completely. Theoretical analysis and experimental results show its effectiveness.
出处
《西安电子科技大学学报》
EI
CAS
CSCD
北大核心
2006年第2期278-280,291,共4页
Journal of Xidian University
基金
国家自然科学基金资助项目(60377034)
关键词
红外焦平面阵列
疵点补偿
运动分析
保持目标边缘
infrared focal plane array
blind pixels compensation
moving analysis
maintain edge
