基于一维平均距离像矩特征的飞机目标识别

Plane Target recognition Based on Average Profile Range Characteristic

本文分析了利用一维距离像进行目标识别的问题。通过对目标相邻的一维距离像进行幅度平方平均处理,得到比较稳定的平均距离像。然后将目标的平均距离像看作二值图像,提取它的矩特征用于分类识别。对三类飞机目标进行分类识别的结果表明该方法是可行的。

The problem of target recognition using radar target one - dimen-tional range profile is analyzed in this article. Through averaging the amplitude square of radar target neighbourly one-dimentional range profiles, the rela-tively stable one -dimentional mean range profile can be obtained. The radar target one-dimentional mean range profile is considered as a 0-1 binary-value image and its moment features are abstracted for classification. The experimental result using these moment features to classify three kinds of airplanes shows this method is feasible.