Onboard locating system using real-time image processing for a self-navigating vehicle

An onboard location system for a self-navigating vehicle using signposts with bar codes placed along the routes is described. The system has two independent measuring subsystems. The first is a machine vision system composed of an onboard camera, an onboard image processing unit, and signposts. This subsystem measures location and direction of the vehicle by triangulation when the signposts are detected. The second is a dead-reckoning subsystem based on rotation counts for the right and left wheels. Data from these two subsystems are systematically combined by an extended Kalman filter. The experimental results show that an accurate and smooth estimation of the location and direction of the vehicle is obtained