一种基于形态点特征的目标识别方法

Target Identification Based on Morphological Point Feature

脉冲激光扫描探测体制激光引信的处理对象一般为目标局部图像,由于获得的图像信息少,而且弹目交会姿态复杂多样,局部图像变化较大,识别比较困难。针对这一问题,提出了一种基于形态点特征的目标识别方法。该方法通过回波处理识别云雾目标,利用面积特征排除鸟类干扰,然后用形态学的处理方法对目标图像矩阵进行处理,提取形态点特征,利用点特征进行目标判别。进行了云雾回波实验并作了回波处理仿真,然后利用三维飞机模型对飞机姿态模拟,用得到的150个多种姿态的局部飞机图像进行了算法的验证,结果表明,92%能够进行识别,对弹目交会姿态适应性较强。

Because most of the image matrixes are parts of targets acquired by the laser fuze of pulse scanning laser detection,and partial image matrixes change greatly for complicated missile-target encounter conditions,using which is difficult to identify the target.In order to solve the problem,an target identifying method based on morphological point feature was put forward in this paper.Echo identification was used to distinguish target from cloud.Area feature was used to distinguish target from birds,and morphological points feature was used to identify target.The three-dimensional model of F-15 fighter aircraft was used to simulate the aircraft status.Then,the method was validated with more than 150 partial plane images.The statistical result was that 92% of image matrixes could be identified.The method have a good adaptability to complicated missile-target encounter conditions.