A Hybrid SLAM method for service robots in Indoor Environment

This paper proposes efficient approach for mobile robots mapping in dynamic indoor environment based on artificial landmarks. Consider, real indoor environment in which nonstructural factors and dynamic objects may be built hybrid model. In this paper, topological map is built based on square root unscented Kalman filter that improving estimation accuracy of conventional SLAM. An efficient strategy is designed to remove dynamic object data and introduces Metric-based ICP algorithm to update the state matrix for compensating for an odometer error. Base on topological map, local metric map is built. The results of simulation and experiment demonstrate that the proposed approach is able to build topological/metric map characterized by higher accuracy and robustness than that of conventional EKF-SLAM.