基于特征谱特征的机场跑道异物分层检测算法

发布时间：2018-06-21 03:05

本文选题：毫米波雷达 + 机场跑道异物　；参考：《电子与信息学报》2017年11期

【摘要】：强杂波背景下的弱静止目标检测是机场跑道异物(Foreign Object Debris,FOD)监测雷达面临的关键问题。该文提出一种基于特征谱特征和最小最大概率机(Minimax Probability Machine,MPM)的FOD分层检测算法,首先利用杂波图恒虚警(Constant False Alarm Rate,CFAR)将雷达录取回波中的背景杂波和FOD回波(包含虚警)区分开,然后提取特征谱特征将在回波域中差异较小的FOD回波和虚警回波转换到区分性更大的特征域,最后利用MPM分类器实现对FOD和虚警的分类,从而达到降低虚警次数的目的。基于实测数据的试验结果表明,所提方法可以获得较好的检测性能。
[Abstract]:Weak static target detection in the background of strong clutter is a key problem for Foreign object Debris-FODO (FODD) surveillance radar on airport runway. In this paper, a novel hierarchical detection algorithm based on feature spectrum and minimum maximum probability machine (MPM) is proposed. Firstly, the background clutter in the radar recording echo is distinguished from the fod echo (including false alarm) by using constant false alarm (CFAR) of clutter graph. Then, the feature spectrum features are extracted, and the FOD echo and false alarm echo, which have less difference in echo domain, are converted to the more discriminative feature domain. Finally, the classification of fod and false alarm is realized by using MPM classifier, so as to reduce the false alarm frequency. The experimental results based on the measured data show that the proposed method can obtain better detection performance.
【作者单位】：电子信息控制重点实验室;
【分类号】：TN957.51

【相似文献】