计算机工程与应用 ›› 2016, Vol. 52 ›› Issue (16): 137-141.

• 模式识别与人工智能 • 上一篇    下一篇

一种快速的加权矢量图像匹配方法

王宪保,潘海飞,章国琼,何文秀,姚明海   

  1. 浙江工业大学 信息工程学院,杭州 310023
  • 出版日期:2016-08-15 发布日期:2016-08-12

Fast weighted vector image matching algorithm

WANG Xianbao, PAN Haifei, ZHANG Guoqiong, HE Wenxiu, YAO Minghai   

  1. College of Information Engineering, Zhejiang University of Technology, Hangzhou 310023, China
  • Online:2016-08-15 Published:2016-08-12

摘要: 为了解决工业生产中产品的快速定位,提出了一种快速的图像匹配方法。首先运用Surf算法提取出图像的特征点,然后运用形状上下文特征进行匹配,得到图像的大致位置,最后依据特征点描述子及其位置特征,通过加权矢量匹配的方式得到图像的精确位置,实现产品的快速定位。该算法稳定性高,计算速度快。实验结果表明,该算法能满足实时的工业生产要求。

关键词: 图像匹配, 加权矢量匹配, 形状上下文, 特征描述子

Abstract: This paper proposes an image matching method for rapid positioning of products in industrial production. Firstly, this paper uses SURF algorithm to extract image feature points, then utilizes shape context feature to match images and gets the approximate position of the image. According to the feature point’s descriptor and its position, accuracy location of image can be found and rapid positioning of the product can be achieved through weighed vector matching at last. Experimental results show that this algorithm is reliable and speedy that can satisfy the real-time requirements for industrial production.

Key words: image matching, weighted vector matching, shape context, feature descriptor