基于聚类算法的多目标快速特征点匹配算法
|
摘要
计算机视觉中的特征点匹配技术具有广泛的应用场景,包括目标跟踪、物体测距以及虚拟现实技术等。传统的特征点匹配技术只能匹配到图像中的一个目标,不适用于所有应用情况。首先介绍了传统ORB算法提取特征点和其目标匹配的效果,然后提出一种基于K-Means聚类算法的多目标特征点匹配算法,将K-Means算法应用到传统的特征点匹配算法,结合Hamming距离和RANSAC算法,使其可以匹配多个目标物。最后实验结果表明,在多目标匹配测试中,多目标匹配算法具有匹配耗时短、匹配准确性高、旋转不变性好等优点。
Feature point matching has been considered as an important and challenging research task in computer vision. It can be widely used in many applications; for example, object tracking, object measurement and virtual reality. However, most feature point matching algorithm can only be used to match single object even if there are many objects in an image. In this paper, we first introduced the traditional ORB algorithm to extract feature points and the effect of object matching. Then we presented a robust approach for multi-objects feature point matching based on K-Means. The K-Means algorithm was applied to the traditional feature point matching algorithm, which combines Hamming distance and RANSAC algorithm so that it can match multiple objects. Finally, experimental results show that the multi-objects matching algorithm has the advantages of short time, high matching accuracy, robustness to rotation and scaling in multi-objects matching test.
Feature point matching has been considered as an important and challenging research task in computer vision. It can be widely used in many applications; for example, object tracking, object measurement and virtual reality. However, most feature point matching algorithm can only be used to match single object even if there are many objects in an image. In this paper, we first introduced the traditional ORB algorithm to extract feature points and the effect of object matching. Then we presented a robust approach for multi-objects feature point matching based on K-Means. The K-Means algorithm was applied to the traditional feature point matching algorithm, which combines Hamming distance and RANSAC algorithm so that it can match multiple objects. Finally, experimental results show that the multi-objects matching algorithm has the advantages of short time, high matching accuracy, robustness to rotation and scaling in multi-objects matching test.
引文
[1] Y Li, et al. Simultaneous facial feature tracking and facial expression recognition[J]. IEEE Transactions on Image Processing, 2013,22(7):2559-2573.
[2] I F Sbalzarini, P Koumoutsakos. Feature point tracking and trajectory analysis for video imaging in cell biology[J]. Journal of structural biology, 2005,151(2):182-195.
[3] S N Sinha, et al. Feature tracking and matching in video using programmable graphics hardware[J]. Machine Vision and Applications, 2011,22(1):207-217.
[4] X Leng, J Liu, Z Xiong. A real-time image matching algorithm for navigation system based on bifurcation extraction[J]. Acta Automatica Sinica, 2007,33(7):678.
[5] D G Lowe. Distinctive image features from scale-invariant keypoints[J]. International journal of computer vision, 2004,60(2): 91-110.
[6] D G Lowe. Object recognition from local scale-invariant features[C]. Computer vision, 1999. The proceedings of the seventh IEEE international conference on. Ieee, 1999-2: 1150-1157.
[7] H Bay, T Tuytelaars, L Van Gool. Surf: Speeded up robust features[J]. Computer vision–ECCV 2006, 2006:404-417.
[8] E Rublee, et al. ORB: An efficient alternative to SIFT or SURF[C]. Computer Vision (ICCV), 2011 IEEE international conference on. IEEE, 2011:2564-2571.
[9] M Calonder, et al. Brief: Binary robust independent elementary features[J]. Computer Vision–ECCV 2010, 2010:778-792.
[10] E Mair, et al. Adaptive and generic corner detection based on the accelerated segment test[C]//European conference on Computer vision. Springer, Berlin, Heidelberg, 2010: 183-196.
[11] J MacQueen. Some methods for classification and analysis of multivariate observations[C]. Proceedings of the fifth Berkeley symposium on mathematical statistics and probability. 1967,1(14):281-297.
[12] D T Pham, S S Dimov, C D Nguyen. Selection of K in K-means clustering[J]. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 2005,219(1):103-119.
[13] M Norouzi, D J Fleet, R R Salakhutdinov. Hamming distance metric learning[C]. Advances in neural information processing systems. 2012:1061-1069.
[14] G H Norton, A Salagean. On the Hamming distance of linear codes over a finite chain ring[J]. IEEE Transactions on Information Theory, 2000,46(3):1060-1067.
[15] H Yang, Y Wang. A LBP-based face recognition method with Hamming distance constraint[C]. Image and Graphics, 2007. ICIG 2007. Fourth International Conference on. IEEE, 2007:645-649.
[16] K Hogrefe, W Ziegler, G Goldenberg. Measuring the formal diversity of hand gestures by their hamming distance[J]. Integrating gestures: The interdisciplinary nature of gesture, 2011-4.
[17] F Gieseke, et al. Buffer kd trees: processing massive nearest neighbor queries on GPUs[C]. International Conference on Machine Learning. 2014:172-180.
[18] A Barclay, H Kaufmann. FT-RANSAC: Towards robust multi-modal homography estimation[C]. Pattern recognition in remote sensing (PRRS), 2014 8th IAPR workshop on. IEEE, 2014: 1-4.
[2] I F Sbalzarini, P Koumoutsakos. Feature point tracking and trajectory analysis for video imaging in cell biology[J]. Journal of structural biology, 2005,151(2):182-195.
[3] S N Sinha, et al. Feature tracking and matching in video using programmable graphics hardware[J]. Machine Vision and Applications, 2011,22(1):207-217.
[4] X Leng, J Liu, Z Xiong. A real-time image matching algorithm for navigation system based on bifurcation extraction[J]. Acta Automatica Sinica, 2007,33(7):678.
[5] D G Lowe. Distinctive image features from scale-invariant keypoints[J]. International journal of computer vision, 2004,60(2): 91-110.
[6] D G Lowe. Object recognition from local scale-invariant features[C]. Computer vision, 1999. The proceedings of the seventh IEEE international conference on. Ieee, 1999-2: 1150-1157.
[7] H Bay, T Tuytelaars, L Van Gool. Surf: Speeded up robust features[J]. Computer vision–ECCV 2006, 2006:404-417.
[8] E Rublee, et al. ORB: An efficient alternative to SIFT or SURF[C]. Computer Vision (ICCV), 2011 IEEE international conference on. IEEE, 2011:2564-2571.
[9] M Calonder, et al. Brief: Binary robust independent elementary features[J]. Computer Vision–ECCV 2010, 2010:778-792.
[10] E Mair, et al. Adaptive and generic corner detection based on the accelerated segment test[C]//European conference on Computer vision. Springer, Berlin, Heidelberg, 2010: 183-196.
[11] J MacQueen. Some methods for classification and analysis of multivariate observations[C]. Proceedings of the fifth Berkeley symposium on mathematical statistics and probability. 1967,1(14):281-297.
[12] D T Pham, S S Dimov, C D Nguyen. Selection of K in K-means clustering[J]. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 2005,219(1):103-119.
[13] M Norouzi, D J Fleet, R R Salakhutdinov. Hamming distance metric learning[C]. Advances in neural information processing systems. 2012:1061-1069.
[14] G H Norton, A Salagean. On the Hamming distance of linear codes over a finite chain ring[J]. IEEE Transactions on Information Theory, 2000,46(3):1060-1067.
[15] H Yang, Y Wang. A LBP-based face recognition method with Hamming distance constraint[C]. Image and Graphics, 2007. ICIG 2007. Fourth International Conference on. IEEE, 2007:645-649.
[16] K Hogrefe, W Ziegler, G Goldenberg. Measuring the formal diversity of hand gestures by their hamming distance[J]. Integrating gestures: The interdisciplinary nature of gesture, 2011-4.
[17] F Gieseke, et al. Buffer kd trees: processing massive nearest neighbor queries on GPUs[C]. International Conference on Machine Learning. 2014:172-180.
[18] A Barclay, H Kaufmann. FT-RANSAC: Towards robust multi-modal homography estimation[C]. Pattern recognition in remote sensing (PRRS), 2014 8th IAPR workshop on. IEEE, 2014: 1-4.