计算机工程 ›› 2011, Vol. 37 ›› Issue (15): 201-202,211.doi: 10.3969/j.issn.1000-3428.2011.15.064

• 图形图像处理 • 上一篇    下一篇

基于边缘特征的亚像素相关匹配跟踪算法

刘 倩,闫宇壮,黄新生   

  1. (国防科学技术大学机电工程与自动化学院,长沙 410073)
  • 收稿日期:2011-01-14 出版日期:2011-08-05 发布日期:2011-08-05
  • 作者简介:刘 倩(1987-),女,硕士研究生,主研方向:图像处理;闫宇壮,博士研究生;黄新生,教授、博士
  • 基金项目:
    国家部委基金资助项目

Subpixel Correlation Matching Tracking Algorithm Based on Edge Feature

LIU Qian, YAN Yu-zhuang, HUANG Xin-sheng   

  1. (College of Mechatronics Engineering and Automation, National University of Defense Technology, Changsha 410073, China)
  • Received:2011-01-14 Online:2011-08-05 Published:2011-08-05

摘要: 基于像素灰度和基于特征的相关匹配跟踪算法跟踪精度不高,只能匹配定位到整像素点。针对该问题,采用相关函数拟合极值法对整像素匹配点进行亚像素定位。设计一种新的实验方法和基于边缘特征的相关匹配跟踪算法。实验结果表明,该算法可以达到0.1级别的亚像素定位精度。

关键词: 亚像素定位, 边缘提取, 边缘相关匹配, 相关函数拟合极值法, 匹配跟踪

Abstract: Correlation matching tracking algorithm based on pixel gray and character is commonly used. The algorithm based on edge feature is a research hotspot because of its advantages for application. However, this kind of algorithm can just locate to integer pixel for its low tracking precision. A method based on the quadratic surface fitting is selected for subpixel matching through analysis and comparison. A new experiment is designed to show that this algorithm gets as high as 0.1 level of pixel location precision.

Key words: subpixel orientation, edge extraction, edge correlation matching, correlation function fitting extreme method, matching tracking

中图分类号: