Driving Assistance Using Computer Vision

Abstract

Collision is a serious issue that needs to be addressed. Many attempts are being made to address the problem of collision. In this paper, we have presented a driving assistant for collision detection. The technique works with an input video from the dashboard camera mounted in the front of the vehicle which captures the lane in front of the vehicle along with objects and other vehicles. The given approach works on utilizing the input to detect the lane of the vehicle, the lane change of the vehicle, and detect other vehicles and objects using advanced computer vision algorithms and data science-based software applications. Using the information detected the idea is to predict the collision in case it happens beforehand. This small window of alert can be useful in the case of real-time driving assistance using the same technologies. This model aims to provide driving assistance for better support to the driver and hence enhance driver safety based on timely warning.