An Intelligent Vision System to Navigate a Robot Without On-Board Sensors

Abstract

This paper presents a smart system to navigate a robot without sensors using a vertical camera and real-time data processing to identify static and dynamic obstacles. To identify the obstacles, image processing techniques are used such as background subtraction and morphological operations in real-time. After detecting the obstacles, the optimal path is determined, and control commands are sent to the robot accordingly. The navigation path of the robot from the start point (robot current location) to the destination point is determined based on graph theory. A graph is created to cover all possible routes in the work area, then A-Star algorithm is utilized to find the optimal path. The optimal path is updated continuously to account for any moving obstacles that could obstruct the robot's path. © 2019 IEEE.