1A2-F09 Monocular Camera based Arbitrary Object Tracking for Four-wheel UGV(Wheeled Robot/Tracked Vehicle(2))
Nobuya Okada, Satoshi Suzuki · The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec) · 2012
In this paper, we get position and velocity of the arbitrary object relative to the vehicle to realize monocular camera based arbitrary object tracking for four-wheel UGV. Firstly, the object is detected in the camera image by matching local features. Secondly, using particle filter combine matching information with other sensor information, and estimate relative position and velocity. Lastly, we implement matching algorithm and particle filter, and verify their accuracy by experiments.