TRAFFIC LIGHT AND SIGNS DETECTION AND CONTROLLING VEHICLES USING DEEP LEARNING

Abstract

in the Autonomous Driving, traffic lights and signs are very important to provide useful information to the autonomous driving such as direction and alerts, we must build a system for autonomous driving that can detect these traffic signals and signs, and perform action based on these detected traffic signals and signs to ensure our autonomous car to be in lane with the traffics in the urban areas. The main objective of this project is to develop image recognition algorithms to recognize Traffic Signs and Signals and then perform action to control car in correspondence to the traffic sign and signal. This report will describe convolutional neural network CNN based object detection used for the recognition of Traffic Signals and signs, CNN can predict different 2D poses of the signs i.e., triangular, square and circle. Our model can easily detect the signal by differentiating the colors of signal in real time and then process the information and tell the vehicle what to do. The main goal and advantage of using this technique is that it provides better recognition of Traffic Signals and Signs. Model of car is used to represent the algorithm. The model can start running, slow, fast and stop by recognizing the Traffic Signals and Signs. We use different hardware components to represent and test our algorithm on model or vehicle.