An obstacle detection method based on machine vision

To develop a quick obstacle detection algorithm for the image-based autonomous vehicle (AV) or computer assisted driving system, this paper presents a real-time method for road barriers recognition and obstacle height calculation. It adopts two similar cameras with different pitch angles, which can perform real-time monitoring for the front area of the vehicle with different coverage areas. The vehicle will move at even speed on a flat road, cameras will sample images at a given rate and the images will be analyzed simultaneously. Focus will be on the virtual window area of the image which can be converted to the actual distance and speed to the vehicle. Counting of the blackened virtual sub-area will be able to differentiate an obstacle with a height with a shadow on the ground. Experiment applying this method proves the effectiveness of obstacle detection.