Autonomous Navigation Robot Based on Real-Time Image Processing

Abstract

The project discusses the design and operation of a simple robot vehicle guided by the track and navigation system made capable by 2D image processing in MATLAB software. The system uses a robust and simple design to automatically track a robot toward a specified path. It does not use GPS or other position sensor system to track vehicle along the path but rather uses visual information run by a software program. The program takes real-time pictures of the vehicle, understanding its direction of projection, and guides to path set by the user. Its contribution can be used for unmanned autodriving/autopiloting of a vehicle in a given path through a monitoring camera or satellite imagery. Its implication consumes least resources of technology as only the imagery monitor system and simple control signal transmission apart from the complex dependency of resources of GPS technology or multiplexed sensor systems. The only requirement being the visibility of the robot from a static camera device, its approach is similar to the functionality of navigation through biological visual based guidance system of humans, so it is a direct, experienced, and simple robust approach.