A new method for recognition of 3D object

Recognition of 3D object is a very important task in computer vision. The traditional 3D object recognition method is using the projective invariant which can be derived from a complicated space geometric model. In this paper, a new simple space geometric model was proposed, which is a combination of four points and a vector, and the relation between the 3D affine invariants and 2D affine invariants was calculated. Employing the relation the object can be recognized from single view without the attitude of object and the parameter of the camera. We apply the method to the real image

[1]  Paul A. Beardsley,et al.  3D Model Acquisition from Extended Image Sequences , 1996, ECCV.

[2]  Rachid Deriche,et al.  Robust Recovery of the Epipolar Geometry for an Uncalibrated Stereo Rig , 1994, ECCV.

[3]  Isaac Weiss,et al.  Model-Based Recognition of 3D Objects from Single Images , 2001, IEEE Trans. Pattern Anal. Mach. Intell..

[4]  Stefan Carlsson.  Relative Positioning from Model Indexing , 1993, BMVC.

[5]  Ronen Basri,et al.  3-D to 2-D recognition with regions , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[6]  Jezekiel Ben-Arie,et al.  Iconic recognition with affine-invariant spectral signatures , 1996, Proceedings of 13th International Conference on Pattern Recognition.