Computer Engineering and Applications ›› 2017, Vol. 53 ›› Issue (17): 202-206.DOI: 10.3778/j.issn.1002-8331.1506-0207

Previous Articles     Next Articles

Image pair selecting and Hash mapping based fast matching for large scale image collection

LIU Liman1, SUN Kun2, XU Haiyang2, HU Huaifei1   

  1. 1. School of Biomedical Engineering, South-Central University for Nationalities, Wuhan 430074, China
    2. School of Automation, Huazhong University of Science and Technology, Wuhan 430074, China
  • Online:2017-09-01 Published:2017-09-12


刘李漫1,孙  琨2,徐海洋2,胡怀飞1   

  1. 1.中南民族大学 生物医学工程学院,武汉 430074
    2.华中科技大学 自动化学院,武汉 430074

Abstract: This paper addresses the heavy image matching cost when building 3D models from large scale images. Since most images in a collection do not overlap, matching such pairs is cumbersome and does not contribute to the reconstruction. In this paper a Hash-based technique is proposed to filter non-overlapping pairs to avoid exhaustive pairwise matching, thus improve matching efficiency. The proposed method consists of four consecutive steps: image Hash description, matching graph initialization, candidate matching graph selection and Hash matching. Experimental results show that the proposed method can greatly speed up the matching stage in 3D reconstruction.

Key words: 3D reconstruction, fast matching, Hash, matching graph

摘要: 针对基于图像进行三维重建技术在使用大规模图像集合进行重建时,需要对图像集合中图像进行两两匹配耗时问题,提出了基于哈希技术对图像构建全局哈希特征的方法,通过过滤掉无效的图像关系对来减少计算时间,极大地提高了大规模图像集合三维重建的匹配计算效率。提出的大规模图像快速哈希匹配算法包括构建图像哈希特征、构建初始匹配图、挑选候选匹配对、哈希匹配几个步骤。实验结果表明该方法能显著地提高三维重建中图像匹配的速度。

关键词: 三维重建, 快速匹配, 哈希, 匹配图