Duals, Invariants, and the Recognition of Smooth Objects from their Occlucing Contours

This paper presents a new geometric relation between a solid bounded by a smooth surface and its silhouette in images formed under weak perspective projection. The relation has the potential to be used for recognizing complex 3-D objects from a single image. Objects are modeled by showing them to a camera without any knowledge of their motion. The main idea is to consider the dual of the 3-D surface and the family of dual curves of the silhouettes over all viewing directions. Occluding contours correspond to planar slices of the dual surface. We introduce an affine-invariant representation of this surface that can constructed from a sequence of images and allows an object to be recognized from arbitrary viewing directions. We illustrate the proposed object representation scheme through synthetic examples and image contours detected in real images.

[1]  Peter J. Giblin,et al.  Epipolar curves on surfaces , 1995, Image Vis. Comput..

[2]  H. Piaggio.  Differential Geometry of Curves and Surfaces , 1952, Nature.

[3]  D. Kriegman,et al.  On recognizing and positioning curved 3D objects from image contours , 1989, [1989] Proceedings. Workshop on Interpretation of 3D Scenes.

[4]  Gérard G. Medioni,et al.  The Challenge of Generic Object Recognition , 1994, Object Representation in Computer Vision.

[5]  P. Giblin,et al.  Curves and singularities : a geometrical introduction to singularity theory , 1992 .

[6]  David A. Forsyth,et al.  Efficient recognition of rotationally symmetric surfaces and straight homogeneous generalized cylinders , 1993, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[7]  David J. Kriegman,et al.  Invariant-based recognition of complex curved 3D objects from image contours , 1995, Proceedings of IEEE International Conference on Computer Vision.

[8]  David J. Kriegman,et al.  On using CAD models to compute the pose of curved 3D objects , 1992, CVGIP Image Underst..

[9]  Jean Ponce,et al.  Finding the limbs and cusps of generalized cylinders , 1987, Proceedings. 1987 IEEE International Conference on Robotics and Automation.

[10]  Michel Dhome,et al.  Inverse perspective transform from zero-curvature curve points application to the localization of some generalized cylinders , 1989, Proceedings CVPR '89: IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[11]  Farzin Mokhtarian,et al.  The renormalized curvature scale space and the evolution properties of planar curves , 1988, Proceedings CVPR '88: The Computer Society Conference on Computer Vision and Pattern Recognition.

[12]  David J. Kriegman,et al.  On Recognizing and Positioning Curved 3-D Objects from Image Contours , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[13]  Olivier D. Faugeras,et al.  Using Extremal Boundaries for 3-D Object Modeling , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[14]  J J Koenderink,et al.  What Does the Occluding Contour Tell Us about Solid Shape? , 1984, Perception.

[15]  Roberto Cipolla,et al.  Motion from the frontier of curved surfaces , 1995, Proceedings of IEEE International Conference on Computer Vision.

[16]  P. Giblin,et al.  Curves and Singularities , 1984 .

[17]  David J. Kriegman,et al.  Hot curves for modelling and recognition of smooth curved 3D objects , 1997, Image Vis. Comput..

[18]  Roger Mohr,et al.  3-d Structure Inference from Image Sequences , 1991, Int. J. Pattern Recognit. Artif. Intell..

[19]  Michel Dhome,et al.  Inverse Perspective Transform Using Zero-Curvature Contour Points: Application to the Localization of Some Generalized Cylinders from a Single View , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[20]  Michel Dhome,et al.  Finding the perspective projection of an axis of revolution , 1991, Pattern Recognit. Lett..

[21]  Marie-Odile Berger,et al.  3D Surface Reconstruction Using Occluding Contours , 1995, International Journal of Computer Vision.