LAUSR

Recently, stereovision has been used in robotics for obtaining information for real-time mapping. Stereovision makes use of two cameras to solve the problems of a lack of discriminative image features and repetitive patterns and to extend the field of view. Stereovision algorithms and techniques are working to create a map to know its surrounding environment. Despite advances in the field of autonomous robotics, several challenges still exist. The majority of these challenges exist owing to the dynamic nature of the navigational environment and the lack of complete knowledge of new, unknown, and complex environments. In this paper, we create a novel, intelligent, and dynamic mapping system for autonomous navigation using stereovision. The proposed system is called intelligent multi-sensor multi-baseline mapping system (I3MS). The main contribution of this paper is the dynamic tuning of the stereovision parameters to address the dynamics of the environment (various light conditions, movement obstacles, and various times of day). The I3MS works in unstructured, unknown, dynamic, and indoor environments. It is based on an in-depth study of the effects of light and distance from obstacles on the accuracy of the 3-D map. It uses a fuzzy-logic system to tune the stereovision parameters to handle dynamic environments. The obtained experimental results illustrate the robustness and accuracy of the proposed system.