一种对空目标分类识别的设计方法

A design method of classification and recognition facing on the air-targets

针对空中目标旋转部件对雷达回波有调制作用的机理,它利用回波时域幅度调制信息的幅度偏差系数作为调制特征,首先利用仿真数据分析了固定翼飞机目标和直升机飞机目标的差异,随后提出了一种最近邻分类器的设计方法,最后利用实测数据对目标进行分类,取得了较高的识别率。该分类器算法复杂度低,计算量小,识别率高,可以用这种方法对两类目标进行分类识别。

Facing on rotating components for aerial target radar with modulation of the mechanism,it uses the rate deviation coefficient as a modulation characteristic. First, it analyzes the differences between fixed-wing aircraft and helicopter aircraft with simulation data; then a nearest neighbor classifier method is proposed to classify using the measured data on the targets, and it is achieved a high recognition rate. The classification algorithm is low complexity, less computation and high recognition rate. We can use this method to classify two types of targets.