A Practical and Efficient Evaluation Function for 3D Model Based Vehicle Matching

Authors: Yuan Zheng

Abstract:

3D model-based vehicle matching provides a new way for vehicle recognition, localization and tracking. Its key is to construct an evaluation function, also called fitness function, to measure the degree of vehicle matching. The existing fitness functions often poorly perform when the clutter and occlusion exist in traffic scenarios. In this paper, we present a practical and efficient fitness function. Unlike the existing evaluation functions, the proposed fitness function is to study the vehicle matching problem from both local and global perspectives, which exploits the pixel gradient information as well as the silhouette information. In view of the discrepancy between 3D vehicle model and real vehicle, a weighting strategy is introduced to differently treat the fitting of the model’s wireframes. Additionally, a normalization operation for the model’s projection is performed to improve the accuracy of the matching. Experimental results on real traffic videos reveal that the proposed fitness function is efficient and robust to the cluttered background and partial occlusion.

Keywords: 3D-2D matching, fitness function, 3D vehicle model, local image gradient, silhouette information.

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1124439

References:

[1] P. David and D. DeMenthon, “Object recognition in high clutter images using line features,” in Computer Vision, 2005. ICCV 2005. Tenth IEEE International Conference on, vol. 2. IEEE, 2005, pp. 1581–1588.
[2] C. Wiedemann, M. Ulrich, and C. Steger, “Recognition and tracking of 3d objects,” Pattern Recognition, pp. 132–141, 2008.
[3] Y. Guo, C. Rao, S. Samarasekera, J. Kim, R. Kumar, and H. Sawhney, “Matching vehicles under large pose transformations using approximate 3d models and piecewise mrf model,” in Computer Vision and Pattern Recognition, 2008. CVPR 2008. IEEE Conference on. IEEE, 2008, pp. 1–8.
[4] S. Nedevschi, S. Bota, and C. Tomiuc, “Stereo-based pedestrian detection for collision-avoidance applications,” Intelligent Transportation Systems, IEEE Transactions on, vol. 10, no. 3, pp. 380–391, 2009.
[5] J. Liebelt and K. Schertler, “Precise registration of 3d models to images by swarming particles,” in Computer Vision and Pattern Recognition, 2007. CVPR’07. IEEE Conference on. IEEE, 2007, pp. 1–8.
[6] B. Rosenhahn, T. Brox, and J. Weickert, “Three-dimensional shape knowledge for joint image segmentation and pose tracking,” International Journal of Computer Vision, vol. 73, no. 3, pp. 243–262, 2007.
[7] S. Dambreville, R. Sandhu, A. Yezzi, and A. Tannenbaum, “Robust 3d pose estimation and efficient 2d region-based segmentation from a 3d shape prior,” in Computer Vision–ECCV 2008. Springer, 2008, pp. 169–182.
[8] C. Reinbacher, M. R¨uther, and H. Bischof, “Pose estimation of known objects by efficient silhouette matching,” in Proceedings of the 2010 20th International Conference on Pattern Recognition. IEEE Computer Society, 2010, pp. 1080–1083.
[9] V. Prisacariu and I. Reid, “Pwp3d: Real-time segmentation and tracking of 3d objects,” International Journal of Computer Vision, pp. 1–20, 2012.
[10] C. Meilhac and C. Nastar, “Robust fitting of 3d cad models to video streams,” in Image Analysis and Processing. Springer, 1997, pp. 661–668.
[11] E. Rosten and T. Drummond, “Fusing points and lines for high performance tracking,” in Computer Vision, 2005. ICCV 2005. Tenth IEEE International Conference on, vol. 2. IEEE, 2005, pp. 1508–1515.
[12] S. Hinterstoisser, S. Benhimane, and N. Navab, “N3m: Natural 3d markers for real-time object detection and pose estimation,” in Computer Vision, 2007. ICCV 2007. IEEE 11th International Conference on. IEEE, 2007, pp. 1–7.
[13] A. Del Bue, “Adaptive metric registration of 3d models to non-rigid image trajectories,” Computer Vision–ECCV 2010, pp. 87–100, 2010.
[14] S. M. Khan, H. Cheng, D. Matthies, and H. Sawhney, “3d model based vehicle classification in aerial imagery,” in Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on. IEEE, 2010, pp. 1681–1687.
[15] E. Marchand, P. Bouthemy, F. Chaumette, V. Moreau et al., “Robust real-time visual tracking using a 2d-3d model-based approach,” in IEEE Int. Conf. on Computer Vision, ICCV’99, vol. 1, 1999, pp. 262–268.
[16] C. Schmaltz, B. Rosenhahn, T. Brox, D. Cremers, J. Weickert, L. Wietzke, and G. Sommer, “Region-based pose tracking,” in Pattern Recognition and Image Analysis. Springer, 2007, pp. 56–63.
[17] S. Messelodi, C. Modena, and M. Zanin, “A computer vision system for the detection and classification of vehicles at urban road intersections,” Pattern Analysis & Applications, vol. 8, no. 1, pp. 17–31, 2005.
[18] N. Buch, J. Orwell, and S. Velastin, “Urban road user detection and classification using 3d wire frame models,” Computer Vision, IET, vol. 4, no. 2, pp. 105–116, 2010.
[19] Z. Zhang, T. Tan, K. Huang, and Y. Wang, “Three-dimensional deformable-model-based localization and recognition of road vehicles,” Image Processing, IEEE Transactions on, vol. 21, no. 1, pp. 1–13, 2012.
[20] H. Kollnig and H. Nagel, “3d pose estimation by directly matching polyhedral models to gray value gradients,” International Journal of Computer Vision, vol. 23, no. 3, pp. 283–302, 1997.
[21] T. Tan and K. Baker, “Efficient image gradient based vehicle localization,” Image Processing, IEEE Transactions on, vol. 9, no. 8, pp. 1343–1356, 2000.
[22] D. Roller, K. Daniilidis, and H. Nagel, “Model-based object tracking in monocular image sequences of road traffic scenes,” International Journal of Computer Vision, vol. 10, no. 3, pp. 257–281, 1993.
[23] M. Haag and H. Nagel, “Combination of edge element and optical flow estimates for 3d-model-based vehicle tracking in traffic image sequences,” International Journal of Computer Vision, vol. 35, no. 3, pp. 295–319, 1999.
[24] J. Lou, T. Tan, W. Hu, H. Yang, and S. Maybank, “3-d model-based vehicle tracking,” Image Processing, IEEE Transactions on, vol. 14, no. 10, pp. 1561–1569, 2005.
[25] M. J. Leotta and J. L. Mundy, “Vehicle surveillance with a generic, adaptive, 3d vehicle model,” Pattern Analysis and Machine Intelligence, IEEE Transactions on, vol. 33, no. 7, pp. 1457–1469, 2011.
[26] B. Yan, S. Wang, Y. Chen, and X. Ding, “Deformable 3-d model based vehicle matching with weighted hausdorff and eda in traffic surveillance,” in Image Analysis and Signal Processing (IASP), 2010 International Conference on. IEEE, 2010, pp. 22–27.
[27] M. Hodlmoser, B. Micusik, M. Y. Liu, M. Pollefeys, and M. Kampel, “Classification and pose estimation of vehicles in videos by 3d modeling within discrete-continuous optimization,” in 3D Imaging, Modeling, Processing, Visualization and Transmission (3DIMPVT), 2012 Second International Conference on. IEEE, 2012, pp. 198–205.
[28] S. Gupte, O. Masoud, R. Martin, and N. Papanikolopoulos, “Detection and classification of vehicles,” Intelligent Transportation Systems, IEEE Transactions on, vol. 3, no. 1, pp. 37–47, 2002.
[29] Y. Tsin, Y. Genc, and V. Ramesh, “Explicit 3d modeling for vehicle monitoring in non-overlapping cameras,” in Advanced Video and Signal Based Surveillance, 2009. AVSS’09. Sixth IEEE International Conference on. IEEE, 2009, pp. 110–115.
[30] Q. Liu, J. Lou, W. Hu, and T. Tan, “Pose evaluation based on bayesian classification error,” in In Proc. of 14th British Machine Vision Conference, 2003, pp. 409–418.
[31] B. Johansson, J. Wiklund, P. Forss´en, and G. Granlund, “Combining shadow detection and simulation for estimation of vehicle size and position,” Pattern Recognition Letters, vol. 30, no. 8, pp. 751–759, 2009.
[32] S. Jayawardena, M. Hutter, and N. Brewer, “Featureless 2d–3d pose estimation by minimising an illumination-invariant loss,” in Image and Vision Computing New Zealand (IVCNZ), 2010 25th International Conference of. IEEE, 2010, pp. 1–8.
[33] T. Hou, S. Wang, and H. Qin, “Vehicle matching and recognition under large variations of pose and illumination,” in Computer Vision and Pattern Recognition Workshops, 2009. CVPR Workshops 2009. IEEE Computer Society Conference on. IEEE, 2009, pp. 24–29.
[34] K. Brisdon, “Hypothesis verification using iconic matching.” Ph.D. dissertation, University of Reading, 1990.
[35] A. Pece and A. Worrall, “Tracking without feature detection,” in Proc. IEEE Int. Workshop Performance Evaluation of Tracking and Surveillance. Citeseer, 2000, pp. 29–37.
[36] M. Hodlmoser, B. Micusik, M. Pollefeys, M.-Y. Liu, and M. Kampel, “Model-based vehicle pose estimation and tracking in videos using random forests,” in 3DTV-Conference, 2013 International Conference on. IEEE, 2013, pp. 430–437.