一种具有强实时性、强鲁棒性的图像匹配算法

被引:15
作者
李兵 [1 ,2 ,3 ,4 ]
刘磊 [5 ]
魏志强 [6 ]
机构
[1] 新奥特(北京)视频技术有限公司
[2] 中国传媒大学计算机学院
[3] 中关村科技园区海淀园企业博士后
[4] 中国科学院自动化研究所
[5] 北京科技大学计算机与通信工程学院
[6] 北京航空航天大学机器人研究所
基金
北京市自然科学基金;
关键词
RIBRIEF; BRIEF; FAST; 索引; 聚类; 实时性;
D O I
10.13328/j.cnki.jos.004450
中图分类号
TP391.41 [];
学科分类号
080203 ;
摘要
针对描述符BRIEF对图像旋转敏感的问题,提出一种改进的描述符RIBRIEF,该描述符具有识别能力强、提取速度快、占用空间小及抗干扰能力强等优点,并具有旋转不变性.经分析,图像匹配算法的实时性较大程度上由特征点数量、匹配点搜索次数及描述符相似度计算复杂度决定,因此提出通过描述符索引与描述符聚类相结合、基于FAST稳定特征点提取和逻辑运算计算相似度等方法提高算法的整体实时性.实验结果表明,与描述符BRIEF及SURF相比较,基于描述符RIBRIEF的图像匹配算法在鲁棒性及实时性方面均具有明显优势.
引用
收藏
页码:1583 / 1592
页数:10
相关论文
共 12 条
  • [1] Multiple target localization at over 100 fps. Taylor S,Drummond T. Proc.of the British Machine Vision Conf . 2009
  • [2] Online learning of patchperspective rectification for efficient object detection. Hinterstoisser S,Benhimane S,Navab N, et al. Proceedings of theConference on Computer Vision and Pattern Recognition . 2008
  • [3] Machine learning for high-speed corner detection. Rosten E,Drummond T. Proceedings of the 9th European Conference on Computer Vision . 2006
  • [4] Fast Keypoint Recognition in Ten Lines of Code. M. Ozuysal,P. Fua,V. Lepetit. IEEE Conf. on Computer Vision and Pattern Recognition . 2007
  • [5] Simultaneous recognition andhomography extraction of local patches with a simple linear classifier. Hinterstoisser S,Benhimane S,Lepetit V, et al. Proceedings of the British Machine Vision Conference . 2008
  • [6] Real-Time Learning of Accurate Patch Rectification. S.Hinterstoisser,O.Kutter,N.Navab,P.Fua,d V.Lepetit. Proc. IEEE Conf. Computer Vision and Pattern Recognition . 2009
  • [7] Brief: Binary robust independent elementaryfeatures. M. Calonder,V. Lepetit,C. Strecha,P. Fua. Computer Vision–ECCV 2010 . 2010
  • [8] Rapid Object Detection using a Boosted Cascade of Simple Features. Viola P,Jones M. Proceedings of IEEE Conference on Computer Vision and Pattern Recognition . 2001
  • [9] "Dominant orientationtemplates for real-time detection of texture-less objects,". Stefan Hinterstoisser,Vincent Lepetit,Slobodan Ilic, et al. Computer Vision andPattern Recognition (CVPR) . 2010
  • [10] Hyperplane Approximation for Template Matching. Frédéric Jurie,Michel Dhome. IEEE Transactions on Pattern Analysis and Machine Intelligence . 2002