海空背景下红外点目标检测算法

摘要

A grayscale morphology-based algorithm has been developed forinfrared point target detection in the backgrounds of sea and sky.The algorithm consists of three major processing steps.Fist,the original sequence image that have been dilated are accoumulated in order to improve SNR.Second,the original sequence image are filtered by morphological operation in order to suppress the target and noise,then to extract the backbounds.Thirdly.it is calculated that the absolute difference of above two outputs and detection is performent by adaptive thresholding.The experimental results for practical infrared image show that the algorithm can detection the point target with SCR of 1.%提出了一种基于灰度形态学的海空背景下的红外点目标检测算法。该算法分三步：第一步，对原始序列图像进行膨胀后累加，以提高信噪比(SNR)；第二步，对原始序列图像进行形态学滤波，去掉噪声和目标以提取出背景；第三步，前两步结果相减，自适应阈值后即可得到点目标。用实际摄取的图像进行实验，结果表明该算法能检测出信杂比(SCR)为1的红外图像。