论文信息 - Obstacle detection from image divergence and deformation

Obstacle detection from image divergence and deformation

In this paper, we investigate a method for identifying obstacles and measuring the threat of collision. The distortion in images is measured by the change in moments, and is exploited to extract the orientation of surfaces and time-to-contact with the objects. The proposed method was tested on real scenes, and it successfully distinguished between obstacles and road markings.

Roberto Cipolla | Jun Sato | R. Cipolla | J. Sato

[1] Chris Harris,et al. Geometry from visual motion , 1993 .

[2] J. Koenderink. Optic flow , 1986, Vision Research.

[3] Volker Graefe,et al. Vision For Intelligent Road Vehicles , 1993, Proceedings of the Intelligent Vehicles '93 Symposium.

[4] Roberto Cipolla,et al. Image registration using multi-scale texture moments , 1994, Image Vis. Comput..

[5] O. Faugeras,et al. A 3D World Model Builder with a Mobile Robot , 1992 .

[6] Takeo Kanade,et al. Vision and Navigation for the Carnegie-Mellon Navlab , 1987 .

[7] Martial Hebert,et al. Vision and navigation for the Carnegie-Mellon Navlab , 1988 .

[8] Jitendra Malik,et al. An integrated stereo-based approach to automatic vehicle guidance , 1995, Proceedings of IEEE International Conference on Computer Vision.

[9] Andrew Blake,et al. Surface Orientation and Time to Contact from Image Divergence and Deformation , 1992, ECCV.

[10] Andrea J. van Doorn,et al. Invariant Properties of the Motion Parallax Field due to the Movement of Rigid Bodies Relative to an Observer , 1975 .