期刊文献+

基于SIFT和RANSAC的特征图像匹配方法 被引量:40

Features Image Matching Approach Based on SIFT and RANSAC
下载PDF
导出
摘要 针对目前普通图像匹配抗干扰能力不强的问题,将尺度不变特征变换(SIFT)和随机采样一致性(RANSAC)算法结合,提出了一种适应性强的图像匹配算法。首先对图像进行SIFT特征提取,利用最优节点优先搜索并计算最近邻特征向量与次最近邻向量间的欧式距离比来加速完成特征点对预匹配。在此基础上引入随机抽样一致性(RANSAC)算法去除不可靠的匹配对。最后根据匹配点对计算出图像间透射变换的参数。实验结果表明:该匹配算法具有尺度、旋转不变性以及一定的仿射不变性、抗干扰性,可以实现目标物体匹配。 In order to improve the ability of anti-interference in the matching of image feature, a new matching method is proposed based on SIFT and RANSAC. Firstly, using SIFT to extract invariant features from images, best bin first research to be performed and computer the Euclidean distance ratio of nearest neighbor feature vector and second nearest neighbor feature vector to achieve the pre-matching. After that, RANSAC robust estimation is performed to eliminate the wrong feature matching. Finally, transformation parameters between matching and template image is computed by reliable matching points. The experimental results indicate that the matching algorithm is invariant to scale and rotation, and a substantial range of affine distortion, addition of noise, and can effectively complete the object matching between images.
出处 《华东理工大学学报(自然科学版)》 CAS CSCD 北大核心 2012年第6期747-751,共5页 Journal of East China University of Science and Technology
  • 相关文献

参考文献9

  • 1Harris C, Stephens M. A combined corner and edge detector [C]//Fourth Alvey Vision Conference. [ s. l. ]: [ s. n. ]. 1988:147-151.
  • 2Lowe D. Distinctive image features from scale-invariant key points [J]. International Journal of Computer Vision, 2004. 60(2) :91-110.
  • 3Fischler M, Bolles R. Random sample consensus: A para- digm for model fitting with applications to image analysis and automated cartography [J]. Communications of the ACM,1981, 24(6):381-395.
  • 4Hu Xuelong, Tang Yingcheng, Zhang Zhenghua . Video ob- ject matching based on sift algorithm[C]// Conference Neural on Networks & Signal Processing. Nanjing, China: IEEE, 2008:412-415.
  • 5Lowe D. Object recognition from local features from scale- invariant features [C]//International Conference on Comput- er Vision. Kerkyra, Greece: IEEE, 1999 :1150-1157.
  • 6Mikolajczyk K,Sehmid C. An affine invariant interest point detector [J ]. European Conference of Computer Vision, 2002, 15(3);128-130.
  • 7Beis J S, Lowe D G. Shape indexing using approximate nea- rest-neighbour search in high-dimensional spaces [ C]//Con- ference on Computer Vision and Pattern Recognition. Wash- ington D C, USA: IEEE Computer Society, 1997: 1000-1006.
  • 8Wang Wei, Hong Jun, Tang Yiping, Image matching for ge omorphic measurement based on sift and RANSAC methods [C]// International Conference on Computer Science and Software Engineering. Wuhan. Hubei: IEEE, 2008: 317- 320.
  • 9黄有群,付裕,马广焜.基于RANSAC算法的柱面全景图拼接方法[J].沈阳工业大学学报,2008,30(4):461-465. 被引量:18

二级参考文献4

共引文献17

同被引文献293

引证文献40

二级引证文献233

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部