A global localization and self-docking method for mobile robot based on feature map

A global localization and self-docking method for mobile robot is presented. The method is composed of two stages: during the off-line stage, SIFT (scale invariant feature transform) algorithm is used and a DD-BBF (double direction best bin first) matching method is presented to implement the 3-D reconstruction of vision features; an ES (evolution strategy) and adaptive re-sampling scheme were applied in RBPF (Rao-Blackwellized particle filter) to implement the mobile robot SLAM (simultaneous localization and mapping). In the on-line stage, the global docking station is recognized through HMM (Hidden Markov Model) based method, he global metric pose and location of the robot are estimated by a RANSAC algorithm; and then an epipole servoing method is presented to dock the robot precisely. Experiment results carried out with a real robot in an indoor environment show the superior performance of the proposed method.