News
About
Projects
Contact
Syllabus
Calendar
Resources

Omni-Directional Vision System for Mobile Robots

Project Type: Sensor interfacing, Pattern Recognition, Hardware integration, Robotics

Advisors: Ricardo Gutierrez-Osuna, Steve Ortiz

Project Description:

The objective of this project is to design and integrate an omni-directional vision system for mobile robot navigation and self-localization. An omnidirectional imaging system consists of a CCD array coupled with a spherical or hyperbolic mirror that generates a 360-degree view of the surroundings. Fish-eye lens could also been used to generate a wide field of view for the camera. It is expected that the system will employ a small and inexpensive camera, such as the CMUcam system.

Major Goals:

  1. Learn about computer vision, omnidirectional imaging and mobile robotics
  2. Design and fabricate a miniature apparatus for omnidirectional imaging (camera, optics, fixture)
  3. Develop image processing software for navigation and self-localization

Deliverables:

  1. Omnidirectional imaging apparatus
  2. Image capture and analysis software
  3. Schematic diagrams, parts list and instructions to assemble the system
  4. Project documentation, as described in the course syllabus
  5. Live demonstration