MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641048917 A) filed by Mani Deep Karumanchi; Namburi Srinivasa Rao; Annepu Sarayu; Gajji Abhiram; Dasari Maneesha; Kuna Satya Venkata Sai Sri Phanindra Pavan; and Bapatla Engineering College, Bapatla, Andhra Pradesh, on April 17, for 'smart ai-based traffic violation detection using cctv and deep learning techniques.'
Inventor(s) include Mani Deep Karumanchi; Namburi Srinivasa Rao; Annepu Sarayu; Gajji Abhiram; Dasari Maneesha; Kuna Satya Venkata Sai Sri Phanindra Pavan; and Bapatla Engineering College.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "There has been a sharp rise in traffic violations on our roads, owing to increased numbers of vehicles in urban areas. Manual traffic law enforcement becomes inefficient, unreliable, and susceptible to human errors due to fatigue and lack of scalability. Therefore, this project presents an AI-based Traffic Violation Detection System that uses the existing CCTV technology and deep learning algorithms to detect violations in real-time and log them automatically. This work applies to a state-of-the-art object detection algorithm, YOLOv8, which can recognize violations, including riding without a helmet, jumping signal, three-way riding, no seat belt, and violation of designated routes. Deep SORT is used to track the objects through all the video frames. The Automatic License Plate Recognition (ALPR) module uses both the license plates detected via YOLOv8 and then identifies the plate number using OCR (EasyOCR). All violations and corresponding images are stored in the SQLite database. A web app created using Streamlit makes it possible for authorities to access violations data."
Disclaimer: Curated by HT Syndication.
