航空反坦克子母弹对地面目标的图像识别

GROUND TARGET IMAGE RECOGNITION OF AVIATION ANTI-TANK CLUSTER WARHEAD

在分析子母弹运动特性的基础上,提出子母弹对地面目标识别的高实时性要求,并提出一种能满足子母弹高实时性要求的快速图像识别方法。该方法首先用形态算子对实时图像进行滤波,然后用改进的基于投影特征的快速图像匹配算法进行图像匹配识别。通过仿真验证,该算法具有快速的图像识别能力,能满足子母弹图像识别的高实时性要求。

Based on the analysis of the flying properties of cluster warhead, a strict real-timeliness demand of cluster warhead to recognize ground targets is introduced, and a fast image recognition method that meets the strict real-timeliness demand of cluster warhead is put for- ward. The method first of all filters the real-time image with form operators; then it uses the improved fast image matching algorithm based on projection characteristics to match and recognize the image. Simulation validates that the algorithm bears fast image recognition ability so that it meets the strict real-timeliness demand of cluster warhead image recognition.