论文部分内容阅读
为解决低信噪比快速运动点目标的检测问题 ,文中提出了一种基于顺序形态滤波的点目标检测算法。首先采用百分位 p =1的顺序形态滤波提取目标的广义内边缘 ,找出少量候选运动目标点 ,然后利用图像流模型建立目标的运动约束方程 ,根据目标像素在图像序列中运动的连续性和轨迹的一致性原则 ,在多帧连续图像中检测出真实的目标运动轨迹。实验结果证明 ,该方法能够快速、可靠检测信噪比小于 3的运动点目标
In order to solve the problem of detecting fast moving point targets with low signal-to-noise ratio (SNR), a point target detection algorithm based on sequential morphological filtering is proposed. Firstly, the generalized inner edge of the target is filtered by the order form filtering of percentile p = 1, a small number of candidate moving target points are found, and then the motion constraint equation of the target is established by using the image flow model. According to the continuity of the target pixel in the image sequence And trajectory consistency principle, in a multi-frame continuous image to detect the real target trajectory. Experimental results show that this method can quickly and reliably detect moving-point targets with signal-to-noise ratio less than 3