Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (7): 192-193.DOI: 10.3778/j.issn.1002-8331.2010.07.059

• 图形、图像、模式识别 • Previous Articles     Next Articles

Method of corner detection by matching point pair and nucleus

SHENG Zun-bing,CUI Xian-yu,CHEN Er-long   

  1. Advanced Manufacturing Technology Center,Harbin Institute of Technology,Harbin 150001,China
  • Received:2008-09-08 Revised:2008-12-10 Online:2010-03-01 Published:2010-03-01
  • Contact: SHENG Zun-bing

点对核值匹配的角点检测

盛遵冰,崔贤玉,陈二龙   

  1. 哈尔滨工业大学 现代生产技术中心,哈尔滨 150001
  • 通讯作者: 盛遵冰

Abstract: SUSAN algorithm is not good at detecting X corner,so a novelty method is proposed.The concept of point pair is introduced in circular template and corner is detected by using the match rules made between point pair and nucleus.The template response formula is given,which enhances the difference of template responses of pixel,and it improves the performance of anti-interference.Experiments show that this method has good performance to detect both generic corner and X corner.

Key words: Smallest Univalue Segment Assimilating Nucleus(SUSAN) algorithm, corner detection, point pair

摘要: 针对SUSAN法检测X型角点能力较弱的问题,提出一种新的角点检测算法。在圆形模板区域内引入点对的概念,通过建立核值与点对的匹配规则来检测角点。给出了一个模板响应公式,加大了各点模板响应值的差别,提高了算法的抗干扰能力。实验结果表明,该算法对复杂的X型角点和普通角点均具有较好的检测效果。

关键词: SUSAN算法, 角点检测, 点对

CLC Number: