Autonomous Mobile Platform

The aim of the research is to develop a control algorithm for mobile platform exploring the unknown environment. Autonomous mobile platform is equipped with a stereo vision camera Bumblebee2, which is the only source of information about the surrounding environment. Obstacle map is built based on the disparity images . To determine the motion path algorithm proposed in [1] was used. Estimation of the platform motion is performed by corner tracking in successive images. The corners are determined by an algorithm [2]. Corners are tracked by cvCalcOpticalFlowPyrLK function from the OpenCV library [3]. Rejection of incorrect solutions is provided by the RANSAC algorithm [4].

[1] C. Y. Lee, „An algorithm for path connection and its applications", IRE Trans. on Electronic Computers, Vol. EC-10, No. 3, 346-365, 1961.
[2] Rosten E., Drummond T."Machine learning for high-speed corner detection", European Conference on Computer Vision: 430-443, 2006.
[3] Jean-Yves Bouguet "Pyramidal Implementation of the Lucas Kanade Feature Tracker Description of the algorithm", http://robots.stanford.edu/cs223b04/algo_tracking.pdf (last visited November 2010).
[4] Fischler M., Bolles R., "Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography", Communcations of the ACM, 24(6): 381--395, 1981.

Category:

Education

Tags:

• Autonomous
• Mobile
• Platform
• Stereo
• Vision