| This thesis presents a new approach to the problem of constructing a depth map from a sequence of monocular images. The approach only requires accurate knowledge of the robot's motion along the focal axis of the moving camera. When the configuration of the camera is known with respect to the mobile platform, this information can be readily obtained by projecting displacement data from a wheel encoder or range sensors onto the focal axis. Instead of directly calculating the depth to the feature points, we first hypothesize that there is a pair of feature points which have the same depth. Based on this hypothesis, the depths are calculated for all pairs of feature points found in the image using the robot's displacement along the focal axis. As the robot moves, the relative location of two points changes in a specific and predictable manner on the image plane if they are actually at the same depth, in other words, if the hypothesis is correct. The motion of each pair of points on the image plane is observed, and if it is consistent with the predicted behavior, the hypothesis is accepted. Accepted pairs create a graph structure which contains depth relations among the feature points. Depth maps obtained at different time steps are integrated in time using a Kalman filtering-based algorithm to obtain a denser depth map. The algorithm is robust against rotational and translational motion noises, and its performance was experimentally demonstrated using a camera mounted on a mobile platform. The numerical stability and the sensitivity of the algorithm to various noise sources are also discussed. The limitations of the algorithm are analyzed and observed through experiments. |
