弹载激光成像雷达距离像的目标提取技术

Target Extraction Base on Range Image from Missile-Borne Imaging LADAR

为了实现对弹载激光成像雷达距离像中不同目标的准确分离和提取,提出了一种基于其点云数据特点的距离像目标提取方法.通过对点云数据进行聚类和类间距离校正,实现了不同目标的准确分离;通过提取各目标边缘点和遮挡补偿,实现了对被遮挡目标的提取;根据需要提取目标的尺寸设置结构元素大小,采用形态学滤波方法实现了特定目标的快速提取.对实测距离像目标数据的处理结果表明,所提出的方法可以实现弹载激光成像雷达距离像目标的准确提取.

A method was proposed for separating and extracting different targets in LADAR range image base on the characteristics of point cloud data obtained from the missile-borne imaging LADAR to solve some special problems in this case. Different targets were separated accurately by clustering points and adjusting nearby classes according to the minimum distance between points. The obscured targets were found by searching the edge points, and the compensation strategy was proposed. The probable targets were extracted rapidly base on the morphology method by adjusting the size of the morphology structure element. The experimental result shows the method can be used to extract targets exactly.