A study on obstacle detection using 3D Hough transform with corner

Collision avoidance is essential element for safe navigation of mobile robot. In mobile robot system, various sensors are used to recognize or detect objects which have possibility of collision against robot during driving. As inescapable fault of each sensor, many researchers have studied sensor fusion and compensation methods to overcome the problems. Most popular sensors used for mobile robot are laser and ultrasonic sensors. However it is difficult for these sensors to find obstacle in case of a slope road. For the slope environment, this paper studies vision method. Vision sensor can detect slope by using randomized Hough transform (RHT). This paper suggests a method of detecting an obstacle in a complicated space with slant using RHT and corner information. This paper proves to compare the result of presence RHT.