MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641076564 A) filed by Cvr College Of Engineering on June 20, 2026, for Ai-Driven Heterogeneous Traffic Anomaly Detection System Using Multi-View Camera Fusion On Roadside Edge Units.
Inventor includes Dr. V. Ramakrishna.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: This invention presents an AI-driven traffic anomaly detection system designed for heterogeneous and non-lane Indian traffic environments. The system employs a parallel multi-stream processing architecture operating four independent detection modules simultaneously, including Collision Detection, Trajectory Analysis, Livestock Intrusion Detection, and Behavior Analysis. Specialized deep learning models process synchronized multi-view camera feeds and validate results through a Consensus Validator to minimize false positives. The system performs geometric consistency checks and temporal validation across multiple viewpoints, achieving high anomaly detection accuracy. Real-time processing at 30 frames per second enables millisecond-scale emergency response on roadside edge devices. The invention detects accidents, wrong-way vehicles, abnormal traffic behavior, and livestock intrusions while supporting five-level alert prioritization for emergency services and traffic authorities. Continuous machine learning improves detection accuracy over time. The system is suitable for highways, toll plazas, and urban roads, enhancing traffic safety, reducing fatalities, and enabling efficient emergency response in Indian traffic conditions.
Disclaimer: Curated by HT Syndication.
