摘要
文章改进了自适应最小均方误差(LMS)算法,将基于历史检测效果的自适应最小二乘(RLS)算法应用于连续帧运动红外弱小目标检测。在已知n-1时刻滤波器抽头权系数的情况下,通过简单的更新,求出n时刻的优化滤波器抽头权系数。计算机仿真实验证明该算法能较大地提高图像信噪比,并以较低的虚警率在2～3帧内较为准确地检测出运动红外弱小目标。
In this paper, the algorithm of self-adaptive LMS is improved. The algorithm which based on historical effect of self-adaptive least-squares is applied in weak and small moving targets detection with consecutive frames. The optimal weight coefficient at the frame ＂n＂ can be calculated through simple upgrade when it is known at the frame of ＂n - 1 ＂. The experimental results show that the algorithm can largely increase the SNR of the images, and the moving targets can be detected within 2 or 3 frames with lower false alarm.
出处
《激光与红外》
CAS
CSCD
北大核心
2007年第7期683-686,共4页
Laser & Infrared
关键词
背景预测
RLS算法
红外弱小目标
目标检测
background prediction
RKS algorithm
infrared weak and small targets
target detection
