基于雷达图像的运动目标形态检测及跟踪技术

运动目标的检测与跟踪在实际中应用广泛。对于低空高速运动目标的检测与跟踪常常面临目标信号微弱和地物干扰严重等问题。基于某型飞机着陆雷达回波图像,提出并采用滑动窗口形态匹配算法对帧差的灰度图像进行目标检测,同时采用二次匹配算法分离出微弱信号目标,去除假目标干扰。之后再结合多帧图像进行航迹关联,去除静止地物等目标干扰,实现运动目标的跟踪。实际测试表明:该处理流程算法简单,处理一帧的时间短,可以检测到极其微弱的目标信号,有效抑制各种干扰,捕捉目标的反应时间短,实现了运动目标的实时检测、准确定位和无丢失跟踪。

Detection and tracking of moving targets using the morphology match in radar images

Moving targets detection and tracking is widely used in practice. Problems such as weak signal and severe interference of stationary objects usually occur in detecting and tracking high-speed low-altitude moving targets. Based on some aircraft landing radar echo images, a sliding window morphology match algorithm was presented to detect objects in a gray image which was processed by using frame difference. A second match algorithm was used  to find targets with a weak signal and remove false target interference.  Making a correlation between the pilot trail and the multi frame image could eliminate similar target interference of stationary objects and thus achieve the goal of tracking a moving target. The results of experiments showed that this processing algorithm is simple and has a short frame processing time. It can detect very weak signals, effectively suppress any interference, and quickly capture targets. This algorithm achieves real-time detection of moving objects, accurate positioning and no tracking loss.