A simple obstacle arrangement detection algorithm for indoor mobile robots

Over the last several years, indoor mobile robot systems not only for industrial use but also for home or office use have been developed. When the robot works in a narrow workspace, it often collides with obstacles. We thought that one of the causes of the collisions was an arrangement of circumambient obstacles. The robot often works in an unknown workspace where there are no maps prepared previously. We propose a simple obstacle arrangement detection algorithm for the workspace map creation. The algorithm uses ultrasonic stereo sonar and a single image sensor. The stereo sonar can measure by the distance detection accuracy that is approximately 13 mm of errors out of 1300 mm and the angular detection accuracy that is approximately a few degrees of errors. The image processing algorithm can measure angles of corners and/or edges of obstacles using a laser line generator as an extra light source.