Vision based simultaneous localization and Mapping using sigmapoint particle filter

One of the most challenging problems in mobile robotic is to localize the robot and simultaneously prepare a map of its environment. In this paper we have used a particle filter to solve SLAM problem in an indoor environment by using stereo vision. We have implemented this method on a mobile robot while navigating in a planar surface among landmarks which were detected by Scale Invariant Feature Transform (SIFT) technique. We have also proposed a novel method based on Sigma Point Particle Filter (SPPF) to solve SLAM problem. Simulation results are compared with the Particle Filter (PF) based SLAM with the same scenario. Finally the experimental results which verify the accuracy and efficiency of our proposed algorithm are presented.