Abstract: Road Assistance for Autonomous Vehicles aims to enhance road safety and driving intelligence by employing computer vision and deep learning techniques to detect and interpret road conditions. The system integrates core functionalities: Traffic Sign Detection and Pothole Detection. Traffic Sign Detection utilizes datasets and deep learning models to accurately identify road signs, even under poor lighting or adverse weather conditions. Speed Limit Sign Board Detection focuses specifically on recognizing speed limit signs, ensuring vehicles adhere to traffic regulations. Meanwhile, Pothole Detection employs object detection techniques to identify road hazards. The system leverages advanced deep learning methodologies such as Convolutional Neural Networks (CNNs). The training data includes public road images optimized using pre-processing techniques to enhance performance across various environmental conditions. Initial tests indicate high accuracy for each module, and their integration offers a comprehensive road monitoring solution. The implementation of optimized deep learning models ensures minimal latency, allowing quick and accurate detection of traffic signs and road hazards. This system is designed to function effectively across diverse environmental conditions, making it robust for urban and rural roadways alike. By bridging the gap between artificial intelligence and vehicular safety, this project contributes to the evolution of smart transportation, fostering a future where autonomous vehicles can navigate roads with increased efficiency and reduced risk.

Keywords: Autonomous Vehicles, Traffic Sign Detection, Pothole Detection, Deep Learning, Convolutional Neural Networks (CNN), Road Safety, Computer Vision

| DOI: 10.17148/IARJSET.2025.125185