Optimization of robot self-localization accuracy by automatic visual landmark selection

Abstract

By observing visual landmarks it is possible to continuously update estimated robot position while the robot is moving. In particular, using a triangulation algorithm based on three landmarks the robot position can be estimated each time three landmarks are observed. This procedure can provide a very accurate estimate, but it is shown to be very sensitive to noise, depending on spatial landmark configuration and relative position between robot and landmarks. However, this noise can be predicted for each landmark triplet, so that we can always choose a triplet which is the less sensitive to noise. An algorithm is then proposed for automatic selection of optimal landmarks. Experiments show the accuracy and robustness of the proposed method maintains the error in the estimated position within a few centimeters.