计算机工程

• 人工智能及识别技术 • 上一篇    下一篇

基于Harris和几何哈希法的目标匹配

周陈龙,胡福乔   

  1. (上海交通大学自动化系系统控制与信息处理教育部重点实验室,上海 200240)
  • 收稿日期:2012-08-30 出版日期:2013-11-15 发布日期:2013-11-13
  • 作者简介:周陈龙(1988-),男,硕士研究生,主研方向:模式识别,图像处理;胡福乔,副教授
  • 基金项目:
    国家自然科学基金资助项目(61175009)

Objects Matching Based on Harris and Geometric Hashing Algorithm

ZHOU Chen-long, HU Fu-qiao   

  1. (Key Laboratory of System Control and Information Processing, Ministry of Education, Department of Automation, Shanghai Jiaotong University, Shanghai 200240, China)
  • Received:2012-08-30 Online:2013-11-15 Published:2013-11-13

摘要: 复杂环境下的目标匹配会受到物体缩放、旋转、遮挡及光强变化等影响,是模式识别领域的一项难题。针对该问题,提出一种基于Harris算法和改进几何哈希法的目标匹配方法。利用Harris角点提取算法检测兴趣点,通过改进的几何哈希法实现多目标匹配。实验结果表明,该方法可实现复杂环境下的目标匹配,提高匹配精度和速度。

关键词: 目标匹配, Harris算法, 几何哈希法, 仿射不变性, 尺度不变特征变换

Abstract: It is a very important but also a tough issue to matching objects in the field of pattern recognition as objects may be influenced by themselves such as scale, rotation, obstacle and also intensity change in the process of recognition. In allusion to this problem, an object matching method based on Harris algorithm and geometric Hash algorithm is proposed. The extraction of interested point features combined with the structure information of geometric Hash algorithm is used. Experimental results show that not only it is possible for this method to match complex objects, but also the accuracy and speed are increased.

Key words: objects matching, Harris algorithm, geometric Hash algorithm, affine invariant, Scale Invariant Feature Transform(SIFT)

中图分类号: