Affiliation:
1. Northeastern University at Qinhuangdao
2. Tianjin University
Abstract
A new pose determination method is presented, in which the ground bound is needed together with target body’s accurate geometric characters. Compared with traditional location method, the proposed way utilizes target’s straight line character to match with its model’s on the ground, by which to represent the variation of object attitude. Location result can be seen on the image. The method makes the location process independent on the attitude estimation with high intelligent and simple system.Experimental results show that the proposed algorithm can locate the target accurately and achieve pose visualization.
Publisher
Trans Tech Publications, Ltd.
Reference8 articles.
1. Wu Jun, Xiao Zhitao. Phase information and contour projection based vehicle localization new algorithm. Journal of Optoelectronics Laser, J. 19 (5) (2008) 686-691.
2. Talluri R, Aggarwal J K. Mobile robot self-location using model-image feature correspondence. IEEE Trans. Robotics and Automation, J. 12 (1) (1996) 63-67.
3. Gennery D B. Tracking known three-dimensional objects. In Proc. Conf. American Association of Artificial Intelligence, C. (1982) 13-17.
4. G Verghese, K L Gale, C D Dyer. Real-time, parallel motion tracking of three dimensional objects from spatiotemporal images. in: V. Kumar, P S Gopalakrishnan and L N Kanal, Parallel Algorithms for Machine Intelligence and Vision, Springer-Verlag, Berlin, Heidelberg, New York, 1990, pp.340-359.
5. Pierre Moreels, Pietro Perona, Evaluation of Features Detectors and Descriptors Based on 3D Objects. Tenth IEEE International Conference on Computer Vision,C. 1 (2005) 800-807.