Vision system of Unmanned Ground Vehicle

This article describes the design of a vision sensing system with application to the urban operation of unmanned ground vehicles. This paper details a procedure for the detection, classification, and localization of lane markings and speed bumps to safely and autonomously navigate an urban roadway. The detection algorithm was developed using National Instruments LabVIEW tool using a CCD camera over 1394 data interface. To demonstrate the algorithmpsilas effectiveness, experimental results were obtained when applying the algorithm in extreme lighting conditions including shadow, noise, and bright scenarios which showed the approach as responsive and robust to such variations.