计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (16): 179-181.

• 图形、图像、模式识别 • 上一篇    下一篇

基于矩与角点特征的飞机目标识别

侯一民,伦向敏   

  1. 东北电力大学 自动化工程学院,吉林 132012
  • 收稿日期:2008-01-07 修回日期:2008-04-14 出版日期:2008-06-01 发布日期:2008-06-01
  • 通讯作者: 侯一民

Airplane object recognition based on moment and corner

HOU Yi-min,LUN Xiang-min   

  1. School of Automation Engineering,Northeast Dianli University,Jilin 132012,China
  • Received:2008-01-07 Revised:2008-04-14 Online:2008-06-01 Published:2008-06-01
  • Contact: HOU Yi-min

摘要: 飞机图像中,目标与背景往往对比较为强烈,但目标的位置、角度和尺度变化也较为剧烈,这给飞机目标的识别带来了一定的困难。提出了一种基于目标矩和角点特征的飞机图像识别方法。首先分割待测图像和模板图像中的目标,将分割结果根据其矩特征进行规格化。之后运用待测图像中目标的角点集与模板图像中目标的边缘点集计算两者相似性,实现识别。实验证明,方法具有计算快捷,适应性强的特点。

关键词: 矩, 角点, 目标识别, 相似性

Abstract: Sometimes,in the airplane images,the object pixels have a clear distinction to those of the background.But the position,angle and scale of the object change actively.It troubles the airplane image object recognition.The paper proposes a recognition method based on the moments and corner characters of the object.Firstly,segments the objects in the template image and the image to be measured.The segmentation results would be standard by the moment characters.Then calculats the comparability between the two objects though the edge points of the template image object and the corners of the object in the image to be measured.The experiments show that the proposed method is an efficiency algorithm and has good adaptability.

Key words: moment, corner, object recognition, comparability