Self-Driving Car Using Artificial Intelligence

Abstract

Self-driving cars have the prospective to transform urban mobility by providing safe, convenient and congestion free transportability. This autonomous vehicle as an application of Artificial Intelligence (AI) have several difficulties like traffic light detection, lane end detection, pedestrian, signs etc. This problem can be overcome by using technologies such as Machine Learning (ML), Deep Learning (DL) and Image Processing. In this paper, author’s propose deep neural network for lane and traffic light detection. The model is trained and assessed using the dataset, which contains the front view image frames and the steering angle data captured when driving on the road. Keeping cars inside the lane is a very important feature of self-driving cars. It learns how to keep the vehicle in lane from human driving data. The paper presents Artificial Intelligence based autonomous navigation and obstacle avoidance of self-driving cars, applied with Deep Learning to a simulated cars in an urban environment.