Optimal Correspondence Model for Image Matching With Multi-order Features
-
摘要: 针对图像匹配问题, 提出了一种图像多阶特征对集的最优匹配模型.图像的多阶特征主要是指一阶、二阶和三阶特征, 分别由单个特征点、特征点之间的边或者连接特征点的三角形来定义.最优匹配模型是一个以图像多阶特征为顶点集的加权二分图, 其优点是权重参数可以直接计算, 并能采用Kuhn-Munkras算法求解最大权对集.实验结果表明, 该模型具有很好的鲁棒性, 对于视频序列图像和涂鸦图像, 即使在存在较大缩放、旋转和仿射变换的情况下, 也能获得比较精确的匹配结果, 其准确度通常优于OpenCV中著名的Flann和BruteForce匹配算法.
-
关键词:
- 图像匹配 /
- 多阶特征 /
- 加权二分图 /
- 最大权对集 /
- Kuhn-Munkras算法
Abstract: An optimal correspondence model was proposed for solving image matching problems with multi-order features. A multi-order feature of an image refers to any of its first-, second- and third-order feature, which was defined by a simple feature point, an edge linking two feature points and a triangle connecting three feature points, respectively. The optimal correspondence model was a weighted bipartite graph with multi-order feature as its vertex. With this model the weight could be directly computed and the solution can be easily obtained by the Kuhn-Munkras algorithm. Results show that the model has good robustness for video sequence and graffiti images. Even with obvious rotation, scale, and affine transformation, it can produce a relatively accurate correspondence result, which is usually better than the famous Flann and BruteForce algorithms in OpenCV. -
-
[1] BERG A, BERG T, MALIK J.Shape matching and object recognition using low distortion correspondence[C]∥Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.San Diego:Institute of Electrical and Electronics Engineers Computer Society, 2005, 1:26-33.
[2] ANAND R, ALAN Y, ERIC M.Convergence properties of the softassign quadratic assignment algorithm[J].Neural Computation, 1999, 11:1455-1474.
[3] LAZEBNIK S, SCHMID C, PONCE J.Beyond bags of features:spatial pyramid matching for recognizing natural scene categories[C]∥Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.New York:Institute of Electrical and Electronics Engineers Computer Society, 2006, 2:2169-2178.
[4] LEORDEANU M, HEBERT M.A spectral technique for correspondence problems using pairwise constraints[C]∥Proceedings of the IEEE International Conference on Computer Vision.Beijing:Institute of Electrical and Electronics Engineers Inc., 2005, 2:1482-1489.
[5] SCHMID C, MOHR R.Local gray value invariants for image retrieval[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997, 19 (5) :530-535.
[6] LOWE D.Distinctive image features from scale-invariant keypoints[J].International Journal of Computer Vision, 2004, 60 (2) :91-110.
[7] ZHANG H, BERG A, MAIRE M, et al.SVM-KNN:discriminative nearest neighbor classification for visual category recognition[C]∥Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition.New York:Institute of Electrical and Electronics Engineers Computer Society, 2006, 2:2126-2136.
[8] FISCHLER M, BOLLES C.Random sample consensus:a paradigm for model fitting with applications to image analysis and automated cartography[J].Communications of ACM, 1981, 24 (6) :381-395.
[9] GOLD S, RANGARAJAN A.A graduated assignment algorithm for graph matching[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 1996, 18 (4) :377-388.
[10] MACIEL J, COSTEIRA J.A global solution to sparse correspondence problems[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2003, 25 (2) :187-199.
[11] COUR T, SRINIVASAN P, SHI J.Balanced graph matching[M].Cambridge:MIT Press, 2007:313-320.
[12] CAETANO T, MCAULEY J, CHENG L, et al.Learning graph matching[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31 (6) :1048-1058.
[13] DUCHENNE O, BACH F, KWEON I, et al.A tensorbased algorithm for high-order graph matching[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33 (12) :2383-2395.
[14] 王年, 周梅菊, 夏杰, 等.基于最小生成树与概率松弛结合的谱匹配算法[J].系统工程与电子技术, 2009, 31 (12) :2958-2962.WANG Nian, ZHOU Mei-ju, XIA Jie, et al.Method for spectral correspondence based on minimum spanning tree combined with probabilistic relaxation[J].Systems Engineering and Electronics, 2009, 31 (12) :2958-2962. (in Chinese) [15] 陈华杰, 冯卫平, 林岳松, 等.基于稀疏方位超图匹配的图像配准算法[J].光电子激光, 2010, 21 (12) :1865-1870.CHEN Hua-jie, FENG Wei-ping, LIN Yue-song, et al.Image registration algorithm based on sparse position hyper graph matching[J].Journal of Optoelectronics Laser, 2010, 21 (12) :1865-1870. (in Chinese) [16] 吴飞, 蔡胜渊, 郭同强, 等.三角形约束下的图像特征点匹配方法[J].计算机辅助设计与图形学学报, 2010, 22 (3) :503-510.WU Fei, CAI Sheng-yuan, GUO Tong-qiang, et al.Image feature points matching approach with triangle constraint[J].Journal of Computer-Aided Design&Computer Graphics, 2010, 22 (3) :503-510. (in Chinese) [17] 汤进, 江波, 罗斌.基于图的直方图及路径相似性的匹配方法[J].计算机辅助设计与图形学学报, 2011, 23 (9) :1481-1489.TANG Jin, JIANG Bo, LUO Bin.Graph matching based on graph histogram and path similarity[J].Journal of Computer-Aided Design&Computer Graphics, 2011, 23 (9) :1481-1489. (in Chinese) [18] REN Zhen-hua, ZHAO Jie-yu.Graph matching based a probabilistic spectral method[C]∥Proceedings—20117th International Conference on Natural Computation.New Jersey:IEEE Computer Society, 2011, 3:1512-1517.
[19] GRIMSON W, LOZANO T.Localizing overlapping parts by searching the interpretation tree[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 9 (4) :469-482.
[20] BAY H, TUYTELAARS T, VAN L, et al.SURF:speeded up robust features[J].Computer Vision and Image Understanding, 2008, 110 (3) :346-359.
[21] KOSOWSHY J J, YUILLE A L.The invisible hand algorithm:solving the assignment problem with statistical physics[J].Neural Networks, 1994, 7:477-490.
[22] BELONGGIE S, MALIK J, PUZICHA J.Shape matching and object recognition using shape contexts[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, 24 (4) :509-521.
[23] ZHAO L.CMU/VASC image database[DB/OL].[2011-09-23].http:∥vasc.ri.cmu.edu//idb/html/motion/house/index.html.
[24] MIKOLAJCZYK K.Affine covariant features[EB/OL].[2011-12-20].http:∥www.robots.ox.ac.uk/-vgg/research/affine/.
计量
- 文章访问数: 19
- HTML全文浏览量: 3
- PDF下载量: 8
下载:
