作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2013, Vol. 39 ›› Issue (6): 214-218. doi: 10.3969/j.issn.1000-3428.2013.06.047

• 人工智能及识别技术 • 上一篇    下一篇

一种基于区域分割的地面飞机目标识别方法

王 威1,2,李元垒3,梁 方1,张连峰3   

  1. (1. 长沙理工大学计算机与通信工程学院,长沙 410114;2. 国防科学技术大学ATR重点实验室,长沙 410073; 3. 海军航空工程学院青岛校区,山东 青岛 266041)
  • 收稿日期:2012-06-27 出版日期:2013-06-15 发布日期:2013-06-14
  • 作者简介:王 威(1974-),男,副教授、博士后,主研方向:目标识别,图像分析;李元垒,副研究员、硕士;梁 方,硕士研究生;张连峰,讲师
  • 基金资助:
    国家部委基金资助项目

A Ground Aircraft Target Recognition Method Based on Region Segmentation

WANG Wei 1,2, LI Yuan-lei 3, LIANG Fang 1, ZHANG Lian-feng 3   

  1. (1. Computer and Communication Engineering Institute, Changsha University of Science & Technology, Changsha 410114, China; 2. ATR Key Laboratory, National University of Defense Technology, Changsha 410073, China; 3. Qingdao Branch, Naval Aeronautical Engineering Institute, Qingdao 266041, China)
  • Received:2012-06-27 Online:2013-06-15 Published:2013-06-14

摘要: 针对遥感图像处理中的目标检测问题,根据具体应用要求,采用改进的图像分割方法对地面机场中的某些大型飞机进行型号识别。分析感兴趣的飞机参数,并在先验信息的基础上确定感兴趣区域(ROI),在待检区域和ROI匹配的基础上提取飞机目标,通过比较结构信息和不变矩信息实现地面飞机目标的检测与识别。实验结果表明,该方法的查全率为88.15%,查准率为80.67%,能达到预期目标。

关键词: 区域分割, 飞机识别, 遥感图像, 感兴趣区域, 不变矩, 先验信息

Abstract: Aiming at the object detection problems in remote sensing image processing, according to the specific application requirements, this paper uses the improved image segmentation method to identify the type of some large aircraft at the ground of airport. This paper analyzes the parameter of interest plane, determines the Region of Interest(ROI) on the basis of prior information, and gets the target plane on the basis of inspection area and ROI matching. It compares the structure information and invariant moment to make recognition on the ground plane target. Experimental results show that the recall rate of this method is 88.15%, the precision rate is 80.67%. It can reach the expected purpose.

Key words: region segmentation, aircraft recognition, remote sensing image, Region of Interest(ROI), invariant moment, prior information

中图分类号: