华南理工大学学报(自然科学版) ›› 2006, Vol. 34 ›› Issue (7): 65-68.

• 计算机科学与技术 • 上一篇    下一篇

基于SUSAN算法的分层快速角点检测

邵泽明 朱剑英 王化明   

  1. 南京航空航天大学 机电学院,江苏 南京 210016
  • 收稿日期:2005-10-17 出版日期:2006-07-25 发布日期:2006-07-25
  • 通信作者: 邵泽明(1975-),男,博士生,主要从事图像处理、机器视觉的研究. E-mail:nj_szm@sohu.com
  • 作者简介:邵泽明(1975-),男,博士生,主要从事图像处理、机器视觉的研究.
  • 基金资助:

    国家自然科学基金资助项目(50275078)

Hierarchical Fast Corner Detection Based on SUSAN Algorithm

Shao Ze-ming  Zhu Jian-ying  Wang Hua-ming   

  1. College of Mechanical and Electrical Engineering,Nanjing Univ.of Aeronautics and Astronautics,Nanjing 210016,Jiangsu,China
  • Received:2005-10-17 Online:2006-07-25 Published:2006-07-25
  • Contact: 邵泽明(1975-),男,博士生,主要从事图像处理、机器视觉的研究. E-mail:nj_szm@sohu.com
  • About author:邵泽明(1975-),男,博士生,主要从事图像处理、机器视觉的研究.
  • Supported by:

    国家自然科学基金资助项目(50275078)

摘要: 分析了SUSAN算法在角点检测中运算速度较慢的原因,并提出基于SUSAN算法的分层快速角点检测算法、该算法根据图像中像素周围图像灰度的相似性和角点的特性,引入提升小波变换理论,采用由粗到细的分层策略,首先对图像进行提升小波变换,找到角点的粗略位置,再用SUSAN算法进行精细查找,准确定位角点.实验结果表明,该算法可较大幅度地提高运算速度,节省运算时间.

关键词: SUSAN算法, 角点检测, 分层, 提升小波

Abstract:

The reason for the low speed of SUSAN algorithm in comer detection is analyzed,and a hierarchical fastcomer detection algorithm based on SUSAN algorithm is proposed. According to the gray similarity around pixels and the comer properties in an image,the theory of lifting wavelet transform and the coarse-to-fine hierarchical strategy are adopted to process the image,with the aim of finding the coarse positions of comers. Then,the comers are finely searched and precisely located by using SUSAN algorithm. Experimental results show that the proposed algorithm mRS fast and saves much runtime.

Key words: SUSAN algorithm, comer detection, hierarchy, lifting wavele