Algorithm for Constrained 3D Object Pose Determination from Single Intensity Images
CHENG Lan1, XING Zhe2, TIAN Yuan3
1.Research office, Anyang Institute of Technology, Anyang 455000 2.Department of Electronic Engineering, Shandong University of Science and Technology, Taian 271000 3.Integrated Information System Research Center, Institute of Automation, Chinese Academy of Sciences, Beijing 100080
Abstract Objects, such as road vehicles, are often constrained on a ground plane. The groundplane constraint significantly reduces the pose redundancy of 2D image and 3D line matching. A method is proposed to locate 3D objects on a known plane from a single calibrated intensity image. Firstly, all possible rotation angles of a given model are computed and clustered to obtain some of the most probable rotation angles. Then all possible positions are calculated and clustered for each rotation angle. The candidate rotation angles and positions are filtered and refined to obtain the final results. The proposed method is capable of coping with moderate occlusion. The effectiveness of the method has been verified by experiments.
[1] Roberts L G. Machine Perception of ThreeDimensional Solids // Tippet T, ed. Optical and ElectroOptical Information Processing. Cambridge, USA: MIT Press, 1965: 157197 [2] Grimson W E L. The Combinatorics of Heuristic Search Termination for Object Recognition in Cluttered Environment. IEEE Trans on Pattern Analysis and Machine Intelligence, 1991, 13(9): 920935 [3] Du L, Sullivan G D, Baker K D. Quantitative Analysis of the Viewpoint Consistency Constraint in ModelBased Vision // Proc of the 4th International Conference on Computer Vision. Berlin, Germany, 1993: 632639 [4] Fan T J, Medioni G, Nevatia R. Recognition 3D Objects Using Surface Descriptions. IEEE Trans on Pattern Analysis and Machine Intelligence, 1989, 11(11): 11401157 [6] Grimson W E L, Huttenlocher D P. On the Sensitivity of Hough Transform for Object Recognition. IEEE Trans on Pattern Analysis and Machine Intelligence, 1990, 12(3): 255274 [7] Wolfson H J. ModelBased Object Recognition by Geometric Hashing // Proc of the 1st European Conference on Computer Vision. Antibes, France, 1990: 526536 [8] Dhome M, Richetin M, Lapreste J T, et al. Determination of the Attitude of 3D Objects from a Single Perspective View. IEEE Trans on Pattern Analysis and Machine Intelligence, 1989, 11(12): 12651278 [9] Perrot C G, Hamey L G C. Object Recognition: A Survey of the Literature. Technical Report, 910065C, Sydney, Australia: Macquarie University. School of Mathematics, Physics, Computing and Electronics, 1991 [10]Tan T N, Sullivan G D, Baker K D. Pose Determination and Recognition of Vehicles in Traffic Scenes // Proc of the 3rd European Conference on Computer Vision. Stockholm,Sweden, 1994: 501506 [11]Maybank S J, Worrall A D, Sullivan G D. Filter for Car Tracking Based on Acceleration and Steering Angle // Proc of the British Machine Vision Conference. Edinburgh, Scotland, 1996: 615624 [12] Jurie F. ModelBased Object Tracking in Cluttered Scenes with Occlusions // Proc of the IEEE/RST International Conference on Intelligent Robots and Systems. Grenoble, France, 1997, Ⅱ: 886892 [13] Tan T N, Sullivan G D, Baker K D. ClosedForm Algorithm for Object Pose and Scale Recovery in Constrained Scenes. Pattern Recognition, 1996, 29(3): 449461 [14] Ma Songde, Zhang Zhengyou. Computer Vision-Theory of Computation and Algorithms. Beijing, China: Science Press, 1998: 5294 (in Chinese) (马颂德,张正友.计算机视觉——计算理论与算法基础.北京:科学出版社, 1998: 5294)
2007, Vol. 20 
