Landmarks Exploration Algorithm For Mobile Robot Indoor Localization Using Vision Sensor

In this paper, a novel Landmarks Exploration Algorithm (LEA) is presented for accurate, reliable, and efficient indoor localization. It operates in two stages. The first stage intends to search for artificial Color-Coded Landmarks (CCLs) and store their locations. Extended Kalman Filter (EKF) is exploited in this stage for continually updating the states of the robot, while the camera, equipped with image processing MATLAB code, is used for the detection of landmarks. The second stage attempts to make the robot find a location such that the distance to the detected CCLs is directly measured using a proximity sensor. At this stage, a trilateration method is applied to localize the robot. This paper also proposes an approach to estimate the heading angle of the robot. These two stages contribute to making the robot reach the target as a final step. Furthermore, the LEA performs localization even that one or two CCLs are detected at the same time and also specific order of CCLs is not required. The LEA is implemented, examined, and evaluated inside the CoppeliaSim environment. The simulation results indicate that the LEA provides the robot with the ability to explore CCLs, achieve an accurate localization, and reach the target.