Proof of Concept Design for Terrain Type Recognition in Urban Environments

Solano R.; Huaman J.; Huamanchahua D. · 2022 2nd International Conference on Robotics, Automation and Artificial Intelligence, RAAI 2022 (2022) · DOI: 10.1109/raai56146.2022.10092963

Resumen

The autonomous navigation systems of terrestrial mobile robots have shortcomings in recognizing their environment when they move. Therefore, in this research, a conceptual design for the autonomous displacement of the Andromina robot based on Orbbec Astra S hardware was developed, which enables the recognition of the terrain in urban environments using the machine learning technique. In this article, a recognition system has been designed using the embedded system and image processing of the terrain in urban environments, in addition to the flowchart that allows us to know the sequence of actions throughout the process of recognizing the nature of the terrain. In the same way, the electromechanical design is proposed as part of the proof of concept, accompanied by the study of various sources and using a few methods to delimit the subject.

Instituciones: Universidad Continental