A low-cost autonomous mobile robot localization using artificial landmarks

Abstract

In this research paper, we present a robust, cost-effective, real-time localization method designed for both indoor and outdoor applications, leveraging a straightforward artificial landmark model. This method utilizes image data from artificial landmarks captured by a standard camera and a memory map. The robot computes its position by employing two landmarks prior to movement initiation and one landmark during motion. The use of a single landmark enhances the calculation speed. To enhance location accuracy, the results are subjected to filtering using a Kalman filter. Depending on the predefined constraints, the robot can autonomously or manually chart its course and navigate along it, achieving an average positional error of 4 cm.