Low cost distance estimation system using low resolution single camera and high radius convex mirrors

This paper proposes a single camera based stereo vision system to determine disparity map and the distance of the object from the camera system. The proposed system uses two convex mirrors with a sufficiently long radius of curvature. These two mirrors help to capture a pair of images of the same scene from two different viewpoints just like a conventional two camera based stereo vision system. The use of single camera makes the calibration and rectification process easier. The identical intensity response of the stereo images improves the accuracy of the disparity map for computing the object distance. Graph cut based algorithm is used to compute the disparity and the object distance from the camera system. The experimental results justify the claims.