为解决靶场红外运动目标检测问题,提出了一种实用的检测方法.首先利用差分图像法对运动敏感的特点,实施运动区域检测,然后采用基于恒虚警率方法确定阈值,实施二值化,并分别向横纵坐标投影,取极大值点的坐标作为目标位置的粗略估计,最后在提取出的目标区域实施精确检测.仿真实验结果表明,该方法目标检测效果较好,可实现像素精度定位,能有效抑制背景,且运算量小,实时性强,适用于多种情况.%In order to solve the detection problem of IR moving targets in shooting range, a detection algorithm based on practicality is proposed. Detected the moving region, based on the characteristic of the difference images that is sensitivity to moving. Then thresholding by using the method of constant false alarm rate(CFAR). Getting the estimative value of the target by the projection of the binarized difference image on the y-axis and x-axis separately. Finally ,doing precision detection from the region that including target. The experimental result shows that the target can be detected well and achieved pixel precision. Removed background in effect. The method is small amount of calculation, realtime, can be used in multi-case.
展开▼