A distributed omnidirectional vision sensor

We describe an omnidirectional vision sensor for vehicle applications such as driver assist systems. The system is composed of four hyperbolic mirrors and uses a single camera to image the 360/spl deg/ scene around the vehicle. This is achieved by having each of the hyperbolic mirrors image about one third of the field of view, and having all four of them sharing one common focal point, where an imaging device is positioned. Several constraints and requirements were taken into consideration in the design of the distributed omnidirectional system. First, the design that we present can practically be mounted on a commercial vehicle. Second, it satisfies - to an approximation - the single viewpoint constraint. Third, it requires a single camera to image the whole field of view. A simulation was also done to plot the field of view as a function of the various system parameters.

[1]  Massimo Bertozzi,et al.  GOLD: a parallel real-time stereo vision system for generic obstacle and lane detection , 1998, IEEE Trans. Image Process..

[2]  M. Srinivasan,et al.  Range estimation with a panoramic visual sensor , 1997 .

[3]  K. Schreiner Night Vision: Infrared Takes to the Road , 1999, IEEE Computer Graphics and Applications.

[4]  S. Tokoro Automotive application systems of a millimeter-wave radar , 1996, Proceedings of Conference on Intelligent Vehicles.

[5]  G.T. Clemence,et al.  The application of acoustic ranging to the automatic control of a ground vehicle , 1982, IEEE Transactions on Vehicular Technology.

[6]  Yasushi Yagi,et al.  Real-time omnidirectional image sensor (COPIS) for vision-guided navigation , 1994, IEEE Trans. Robotics Autom..

[7]  Thad Starner,et al.  Finding location using omnidirectional video on a wearable computing platform , 2000, Digest of Papers. Fourth International Symposium on Wearable Computers.

[8]  H. Bunke,et al.  VISION-BASED DRIVER ASSISTANCE USING RANGE IMAGERY , 1998 .

[9]  Sridhar Lakshmanan,et al.  CLARK: a heterogeneous sensor fusion method for finding lanes and obstacles , 2000, Image Vis. Comput..

[10]  Yasushi Yagi,et al.  Obstacle detection with omnidirectional image sensor HyperOmni Vision , 1995, Proceedings of 1995 IEEE International Conference on Robotics and Automation.

[11]  M. Donath,et al.  A virtual mirror for assisting drivers , 2000, Proceedings of the IEEE Intelligent Vehicles Symposium 2000 (Cat. No.00TH8511).