The paper presents a algorithm of visual simultaneous localization and mapping (vSLAM) for a small-size humanoid robot. The algorithm includes the procedures of image feature detection, good feature selection, image depth calculation, and feature state estimation. To ensure robust feature detection and tracking, the procedure is improved by utilizing the method of Speeded Up Robust Features (SURF). Meanwhile, the procedures of image depth calculation and state estimation are integrated in an extended Kalman filter (EKF) based estimation algorithm. All the computation schemes of the visual SLAM are implemented on a small-size humanoid robot with low-cost Window-based PC. Experimentation is performed and the results show that the performance of the proposed algorithm is efficient for robot visual SLAM in the environments.
Tamkang Journal of Science and Engineering=淡江理工學刊 14(2), pp.123-129