Κατασκευή αυτοκινούμενου οχήματος προορισμένου για οδήγηση σε οδούς χωρίς διαγράμμιση αξιοποιώντας λαμβανόμενη εικόνα και τεχνητό νευρωνικό δίκτυο για την ανάλυσή της

Abstract

This thesis ‘s subject is the planning and construction of a vehicle able to autonomously navigate road that is successfully recognized via optical means, analyzed by an artificial neural network constituted of neurons with Radial Basis activation Function (RBF). The vehicle is designed for use in real environments where it attempts to perceive the road as a driver would, by seeing, understanding where the drivable road is and leading the vehicle to the correct direction. Like so the built project will drive to the corresponding direction after analyzing the environment. The Computer Vision OpenCV library is extensively used throughout the project for formatting and making the received motion picture suitable for analysis.