针对飞机目标的分类问题 ,介绍了一种基于局部双谱的高分辨雷达目标距离像识别方法 ,利用 Fisher类别可分性鉴别测度在距离像的双谱平面选择用于分类的局部双谱。将局部双谱方法与现有的几种基于双谱的方法进行比较 ,分析出局部双谱既利用了双谱的平移不变性 ,实现了一维匹配 ,又避免了在识别过程中计算整个双谱 ,减少了计算量 ,而且不需考虑极坐标双谱的插值与积分步长问题。利用三种缩比模型飞机的微波暗室转台数据对局部双谱的分类性能进行了测试 ,结果表明局部双谱是一种很有价值的平移不变特征 ,对于高分辨雷达目标距离像具有较高的识别率 Aiming at the classification problem of aircraft targets, a method of identifying target distance of high resolution radar based on local bispectrum is introduced. The bispectrum plane of distance image is selected for classification of local bispectrum using Fisher classification separability discrimination measure. By comparing the local bispectrum method with several existing bispectrum-based methods, it is analyzed that the local bispectrum not only utilizes the translational invariance of the bispectrum but also realizes the one-dimensional matching and avoids the calculation of the whole double Spectrum, reducing the amount of computation, and without having to consider the interpolation and integral steps of polar bispectrum. The classification performance of the local bispectrum was tested by using microwave anechoic chamber turntable data of three reduced-ratio aircraft models. The results show that the local bispectrum is a valuable translational invariant feature and has a high value for high resolution radar target range images Recognition rate