MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641100536 A) filed by Manivannan K; V. Kavitha; S. Nandhini Devi; R. Muthchelvan; and V. S. B. Engineering College, Karur on August 20, 2026, for Smart Road Accident Detection And Automated Emergency Notification Genai-Enhanced Computer Vision System.
Inventors include Manivannan K; V. Kavitha; S. Nandhini Devi; R. Muthchelvan; and V. S. B. Engineering College, Karur.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: The proposed system presents a Gen-AI-based computer vision system for real-time road accident detection and automated emergency notification. The system uses YOLO-based object detection to identify and track vehicles, pedestrians, and possible collision events from CCTV, dashcam, or traffic-camera video. Deep-learning-based behavioural analysis examines vehicle speed variation, path deviation, inter-vehicle distance, collision patterns, and other spatio-temporal characteristics to distinguish actual accidents from normal traffic events. After an accident is detected, the system determines its severity as minor, moderate, or critical. The Generative AI component converts the detected information into a structured accident report containing relevant details such as accident characteristics, severity, location, and timestamp. The system automatically generates emergency notifications and transmits the information to appropriate emergency authorities, including ambulance, police, and hospital services. The proposed framework aims to reduce human dependency, improve emergency response time, support intelligent decision-making, and enhance road safety in smart-city environments.
Disclaimer: Curated by HT Syndication.
