DriveAlert: A Smart Driving Assistance
Albin A P, Krishnapriya U, Nayana Noshy, Yadhukrishnan R, Krishnaveni V V
Dept. of CSE, College of Engineering Kidangoor, Kottayam, Kerala, India
albinap952654@gmail.com, krishnapriyau03@gmail.com,, noshynayana@gmail.com, iyadhukrishnanr@gmail.com, veni.anuraj@gmail.com
Abstract——This Intelligent Driver Assistance System enhances road safety by integrating real-time signboard detection, speed monitoring, and 3D map visualization. Utilizing a YOLO-based deep learning model, it processes images from a vehicle’s front-facing camera to identify traffic speed signs, with a FastAPI-based backend handling detection and returning real-time results. The Flutter-based front-end provides an intuitive interface with visual and auditory alerts for detected signs and speed limits, ensuring drivers receive timely notifications. Additionally, the system features an over-speed monitoring mechanism that warns drivers if they exceed speed limits, promoting safer driving. To enhance spatial awareness, a 3D map visualization powered by Mapbox offers custom styles for improved navigation and situational awareness. Unlike high-end Ad- vanced Driver Assistance Systems (ADAS) that require expensive hardware, this project delivers a cost-effective solution by leveraging machine learning, computer vision, and interactive UI components, making it accessible for a wider range of vehicles. By offering real-time detection, intelligent notifications, and enhanced road visibility, this system significantly contributes to reducing road accidents and improving driver assistance.
Index Terms—Driver assistance, road safety, sign- board detection, speed monitoring, object recognition, YOLOv8, machine learning, computer vision, real-time alerts, FastAPI, Flutter, mobile application, 3D mapping, Mapbox, over-speed warning, visual alerts, auditory alerts, intelligent driving, smart navigation, traffic sign recogni- tion, driving assistance, road hazard detection, accident prevention, cost-effective ADAS.