计算机工程与应用 ›› 2013, Vol. 49 ›› Issue (4): 174-177.

• 图形图像处理 • 上一篇    下一篇

基于结构特征的飞机目标识别方法

才让卓玛1,张名成2   

  1. 1.西北民族大学 电气工程学院,兰州 730124
    2.酒泉卫星发射中心,甘肃 酒泉 732750
  • 出版日期:2013-02-15 发布日期:2013-02-18

Aircraft recognition via structural features

Cairangzhuoma1, ZHANG Mingcheng2   

  1. 1.School of Electrical Engineering, Northwest University for Nationalities, Lanzhou 730124, China
    2.Jiuquan Satellite Launch Center, Jiuquan, Gansu 732750, China
  • Online:2013-02-15 Published:2013-02-18

摘要: 针对高分辨率可见光图像的飞机目标智能识别,提出一种基于飞机结构特征的识别方案。采用Mean shift方法分割图像,提取区域边界,得到目标轮廓集,对轮廓采取多种结构化表示方式,便于分析轮廓的特征及相互关系;综合利用尺寸、占空比、长宽比、凸包、直线段等多个形状特征筛选轮廓,定位飞机目标,确定飞机轴向,获取目标几何参数。实验结果证明,该方法定位精确、识别率高,是一种快速、稳健的飞机识别方案,对类似的人工目标的识别具有一定的参考意义。

关键词: 自动目标识别, 飞机, 结构特征, 凸包

Abstract: Auto recognition of aircraft in high resolution visual-band image is difficult because of clutter, shadows, self-occlusion and poor image quality. In this paper, a solution is presented based on geometric features. The image is segmented with mean shift algorithm which splits the image into some fields, and then contour of each field is traced and grouped hierarchically. According to structural features of aircraft, some geometric features are used to determine the target contour. Experimental results demonstrating the effectiveness and performance are given. It locates aircraft target with a good precision and efficiency.

Key words: automatic target recognition, aircraft, structural feature, convex hull