摘要
红外焦平面阵列(IRFPA)器件普遍存在着响应度的非均匀性问题,而且这种非均匀性随时间和环境改变会发生缓慢变化,目前常用的一次性校正算法不能适应这种变化。同时由于图像场景的多样性,现有的统计校正算法也存在着一定的适应性问题。本文结合图像运动分析中的光流技术,提出了一种新的基于场景的连续校正算法,该算法具有较好的自适应性能。
The non-uniformity problem is widely existed in IR focal plane arrays (IRFPA). This non-uniformity will change slowly with time and environment conditions. The common disposable correction algorithm cannot suit this change. At the same time, since the diversity in image scene, the existing statistic correction algorithm still existed some adaptability problems. Based on the optical flow technique in image motion analysis, a new continuous correction algorithm based on scene is proposed in the paper. The algorithm has a better adaptive performance.
出处
《光电工程》
CAS
CSCD
北大核心
2001年第5期40-42,共3页
Opto-Electronic Engineering
基金
国家自然科学基金(69677010)资助项目
关键词
红外焦平面列阵
非均匀性校正
自适应校正算法
光流
Infrared focal plane arrays
Non-uniformity correction
Adaptive correction algorithm
Optical flow
