中文摘要：

针对红外成像与跟踪系统中舰船目标自动识别转跟踪后的跟踪点自动选取问题,提出了一种新的舰船目标要害点自动检测算法。基于舰船吃水线位于船体位置下方的几何关系以及动力舱的高亮度特征,检测吃水线和动力舱的交叉部位作为首要攻击点,将吃水线中部作为次要攻击点。首先对目标区域进行去噪预处理,通过Canny边缘提取、边缘闭合、两次连通域去噪等操作滤除背景纹理杂波确定目标的边缘,然后对目标底部边缘进行分段直线拟合,将拟合结果作为吃水线,最后利用灰度值统计确定最亮区域为求取的动力舱位置。实验结果统计表明,该方法检测结果稳定,具有良好的鲁棒性,适用于不同观测方向下的红外舰船目标的要害点检测,为跟踪点的自动选取提供了有效解决方案。

英文摘要：

In order to solve the problem of automatic selection of tracking point in target tracking after automatic target recognition of IR warship targets in the infrared imaging and tracking systems, a new automatic detection algorithm of ship aimpoint was proposed. Based on the waterline′s geometrical relationship below the warship and the high brightness characteristics of engine compartment, the cross part of the waterline and engine compartment was detected. This cross part was taken as the primary aimpoint of IR warship target. Otherwise, the central part of the waterline would be taken as the aim point of IR warship target. First of all, the de-noising pretreatment of target area was used, through the Canny egde-detection, edge closed and two connected domain de-noising operations, the background texture clutter was removed to determine the edge of target. Then, piecewise linear fitting on the target bottom edge was used and the curve fitting results would be taken as the waterline. Finally, the gray value statistics was calculated to get the brightest area as the position of engine compartment. The experimental results show that this method has stability testing results and good robustness. It can be applied to detect the aim point of IR warship target in different viewing directions. The algorithm provides an effective solution for the problem of automatic selection of tracking point in target tracking.