Pose estimation and recognition of ground vehicles in aerial reconnaissance imagery

A surface-based mutual information metric has been proposed by Viola and Wells (1995) to orientate 3D models accurately with 2D imagery in the medical field. In this paper, their objective function is applied and extended to estimate the pose and obtain generic recognition of vehicles in military reconnaissance imagery using appropriate 3D CAD models. Successful orientation and recognition of imaged vehicles has been achieved through the addition of another term to the metric. The new metric term utilises edge information and provides a greater tolerance of initial model translational and rotational offsets.

[1]  Paul A. Viola.  Alignment by maximisation of mutual information , 1993 .

[2]  Paul A. Viola,et al.  Alignment by Maximization of Mutual Information , 1995, Proceedings of IEEE International Conference on Computer Vision.

[3]  Takeo Kanade,et al.  Real-time 3-D pose estimation using a high-speed range sensor , 1993, Proceedings of the 1994 IEEE International Conference on Robotics and Automation.

[4]  Athanasios Papoulis,et al.  Probability, Random Variables and Stochastic Processes , 1965 .

[5]  Paul J. Besl,et al.  A Method for Registration of 3-D Shapes , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Christoph Zierl,et al.  Robust Video-Based Object Recognition Using CAD Models , 1995 .

[7]  Gerd Hirzinger,et al.  Registration of CAD-models to images by iterative inverse perspective matching , 1996, Proceedings of 13th International Conference on Pattern Recognition.

[8]  Yoshihiko Nomura,et al.  3-D object pose estimation based on iterative image matching: Shading and edge data fusion , 1996, Proceedings of 13th International Conference on Pattern Recognition.