Angle Detection of Steering in a Self-Driving Car

Abstract

An exciting new era that has come into focus is the age of self-learning and automation, where one of its application is the self-driven vehicles. In present era, humans have to no longer care about the obstacles in the driving paths or the stressful rush hour traffics, since the automated vehicles can help reach the destination fast and efficiently. Udacity has provided a dataset containing a set of images with steering wheel angle recorded during driving. With the help of the dataset provided, prediction of the steering wheel rotation angle can be done. In order to predict the angle, computer vision is used. It is the main technology facilitating self-driven vehicles. In this project, a system is created to drive a vehicle automatically without any human input requirement. ML models like CNN and OpenCV are used for certain application to detect an object, vehicle, traffic sign, etc. The project is deployed as a simulation software that provides the output in a frame. There are two frames: one that consists the steering wheel and other that consists the road.