Autonomous obstacle avoidance using computer vision for electric auto

Abstract:

Electric Vehicles (EVs) are rapidly transforming the global transportation landscape due to their environmental advantages, improved energy efficiency, and reduced dependence on fossil fuels. However, as EV adoption increases, ensuring vehicle safety through intelligent automation becomes increasingly important. Advanced Driver Assistance Systems (ADAS) play a critical role in improving vehicle safety by assisting drivers in detecting obstacles, maintaining proper lane discipline, and preventing potential accidents. This paper presents EV AUTO, a vision-based prototype designed to perform autonomous obstacle avoidance functions for electric vehicles. Unlike traditional safety systems that depend on expensive sensors such as LiDAR or basic proximity sensors like ultrasonic modules, the proposed system uses computer vision and deep learning techniques to interpret the surrounding road environment in real time. The system employs OpenCV for image processing and a Convolutional Neural Network (CNN) model to detect lane boundaries and identify obstacles such as vehicles, pedestrians, or road hazards. The camera module captures continuous road images, which are processed using image preprocessing techniques including noise reduction, edge detection, and feature extraction. The CNN model analyzes these processed frames to identify lane markings and classify obstacles in the vehicle's path. Based on the detected information, the system automatically generates control commands that adjust the vehicle’s steering and speed to maintain lane alignment and avoid potential collisions. Simulation-based experiments demonstrate that vision-based automation significantly improves system reliability and decision-making capability compared to rule-based sensor systems. Furthermore, the proposed solution maintains a low-cost architecture, making it suitable for integration into mass-market electric vehicles rather than being restricted to high-end automotive platforms. The EV AUTO system highlights how artificial intelligence, computer vision, and embedded systems can work together to create intelligent safety solutions for next-generation electric mobility.