Single Cone Mirror Omni-Directional Stereo

Omni-directional view and stereo information for scene points are both crucial in many computer vision applications. In some demanding applications like autonomous robots, we need to acquire both in real-time without sacrificing too much image resolution. This work describes a novel method to meet all the stringent demands with relatively simple setup and off-the-shelf equipments. Only one simple reflective surface and two regular (perspective) camera views are needed. First we describe the novel stereo method. Then we discuss some variations in practical implementation and their respective tradeoffs. Comments University of Pennsylvania Department of Computer and Information Science Technical Report No. MSCIS-01-03. This technical report is available at ScholarlyCommons: http://repository.upenn.edu/cis_reports/141 Single Cone Mirror Omni-Directional Stereo Shih-Schoii Lin, Ruzena Bajcsy Technical Report MS-CIS-0 1-03 GRASP Laboratory, Computer and Information Science Department University of Pennsylvania, shschon@,qrasp.cis.upenn. edu , rbaicsy@nsfgov (also, bajcsy@grasp.cis.upenn.edu)

[1]  Ryad Benosman,et al.  Multidirectional stereovision sensor, calibration and scenes reconstruction , 1996, Proceedings of 13th International Conference on Pattern Recognition.

[2]  Anup Basu,et al.  Panoramic stereo , 1996, Proceedings of 13th International Conference on Pattern Recognition.

[3]  Ruzena Bajcsy,et al.  The True Single View Point ( SVP ) Configuration for Omni-Directional View Catadioptric System Using Cone Mirror , 2001 .

[4]  Naokazu Yokoya,et al.  Generation of high-resolution stereo panoramic images by omnidirectional imaging sensor using hexagonal pyramidal mirrors , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[5]  Shree K. Nayar,et al.  Rectified catadioptric stereo sensors , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[6]  Hiroshi Ishiguro,et al.  Omni-directional stereo for making global map , 1990, [1990] Proceedings Third International Conference on Computer Vision.

[7]  Shree K. Nayar,et al.  Stereo with mirrors , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[8]  David W. Murray.  Recovering Range Using Virtual Multicamera Stereo , 1995, Comput. Vis. Image Underst..

[9]  Karen B. Sarachik.  Characterising an indoor environment with a mobile robot and uncalibrated stereo , 1989, Proceedings, 1989 International Conference on Robotics and Automation.