基于雷达图像的飞鸟目标检测与信息提取

针对鸟击问题,基于一种自行研制的雷达探鸟实验系统获取了50帧含有飞鸟目标的雷达PPI图像,该系统由X波段海事雷达改造而成.经过背景差分、图像平滑、阈值分割和信息提取4个步骤,将飞鸟目标从复杂的PPI图像中分离出来,提取出包括目标质心位置、目标平均灰度、目标面积、目标形状在内的鸟情信息,实现了一种可行的飞鸟目标检测与信息提取算法.本算法为建立实用的机场雷达探鸟系统奠定了重要的理论基础.

Flying bird targets detection and information extraction based on radar images

To study the bird strike problem,50 radar plane position indicator(PPI) images containing flying bird targets are obtained using a self-made avian radar experimental system based on an adapted X-band marine radar.Flying bird targets are separated from complicated PPI images by four steps: background subtraction,image smoothing,threshold segmentation and information extraction.Information including mass centre,average gray scale value,area and the shape of the avian target is extracted,realizing a feasible algorithm for flying bird targets detection and information extraction.The proposed algorithm makes an important theoretic foundation for building an operational airport-based avian radar surveillance system.