Proceedings of The 2011 IAJC-ASEE International Conference ISBN 978-1-60643-379-9 Paper 080, INT 303 Centralized Vision-based Controller for Multi Unmanned Guided Vehicle Detection Ravindra Thamma Central Connecticut State University thammarav@ccsu.edu Leela Mohan Kesireddy Central Connecticut State University thammarav@ccsu.edu Haoyu Wang Central Connecticut State University wanghao@ccsu.edu Abstract The traditional track-based unmanned ground vehicle (UGV) cannot deviate from the pre- recognized route due to track limitation in its navigation methods. Track limitation has to be overcome in order to render a UGV more flexible. It is desirable to have a UGV move without predetermined tracks and have the ability to deviate to and from routine routes, in order to have flexibility in tasks. This research proposes a navigation system to aid multiple UGVs in navigating to various locations without any physical tracks and without collision with each other. This research demonstrates the image-recognition-based trackless navigation system to enhance the flexibility of multiple UGVs. This research implements an image recognition algorithm to identify the position and orientation of multiple UGVs using a Centralized Image-based Controller Unit (CIBCU). This CIBCU is connected to a vision system and radio frequency (RF) communicator. The CIBCU implements the image recognition algorithm, anti-collision, navigation algorithm, and centralized control center to track and navigate multiple UGVs without physical tracks. A prototype has been developed to demonstrate and test the Vision-based Navigation System. Statistical analyses have been carried out on this newly developed system to find behavior of positioning error. Introduction Conventionally, controlling an unmanned ground vehicle (UGV) and automatic guided vehicle (AGV) has been a challenge. Tracks serve as a key element for navigation systems for a mobile UGV. UGV tracking is a critical component for providing positions, directions, and travel information for motion along a trajectory with minimal deviation. Many