摘要
提出一种基于数学形态学的远距离红外小目标自动检测的新方法，构造了由几种形态学运算组成的非线性算子。结合刚性小平面的三维分析技术，可快速检测红外图象序列中静止或运动的小目标。经实验证明，与传统的小目标检测算法相比，该方法具有速度快、抗噪性好等方面的优越性。
An automatic detection of small infrared target based on mathematic morphology in long range image sequences is proposed. A nonlinear operator consisting of several morphological operations is constructed, together with the 3-D motion analysis of planar patch and the algorithm can detect static or moving small point target as well. Experimental results show that the algorithm is more satisfactory than traditional methods, the proposed scheme has higher speed and better noise immunity.
出处
《数据采集与处理》
CSCD
1997年第4期276-279,共4页
Journal of Data Acquisition and Processing
基金
航天科学基金
关键词
目标检测
数学形态学
图象处理
图象识别
morphology
target detection
infrared image sequence
planar patch
