In order to improve identification probability of terminal sensitive sub-ammunition under complex battle circumstances, the method of extracting the armor target based on the linear array LiDAR was proposed. Combined with stable scanning theory, 3D point cloud imaging on the scanning area was realized. Firstly, the height and gradient combination threshold segmentation algorithm was proposed by analyzing the height data in the point cloud; Then the coordinate transformation was used to adjust the slope ground, and the seed were automatically acquired through the geometric size of the typical armored targets for regional growth segmentation; finally, the extraction of armored targets was achieved through the geometric features of the target. The simulation shows that linear array LiDAR can accurately extract the armor target at the height of 50-120 m in battlefield, which provides technical support for the target detection of new terminal sensitive sub-ammunition.