MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641110620 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on September 15, 2026, for An Ai-Based Intelligent Road Surface Monitoring System Using Yolov8 Segmentation And Edge Computing.
Inventor includes Mrs. N. Haritha.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: ABSTRACT [0020] The invention provides a real-time deep learning-based system for detecting potholes and road lanes, optimized for operation on an NVIDIA Jetson Nano edge computing platform. Traditional computer vision techniques such as Canny Edge Detection proved insufficient due to sensitivity to lighting changes and limited generalization. The system therefore employs a YOLOv8 segmentation model trained on annotated road images to deliver accurate and robust detection under varied real-world conditions. A modular client-server architecture is used in which a server running inside a Docker container hosts the trained model and processes incoming frames, while a client captures video from a USB camera or pre-recorded sources, transmits frames for inference, and displays the resulting detections with bounding boxes and labels. Hardware and software challenges including power stability, camera access within Docker, and GPU visibility were systematically resolved. The resulting low-latency pipeline supports continuous road monitoring and is suitable for intelligent transportation systems, autonomous vehicle assistance, and smart- city infrastructure applications.
Disclaimer: Curated by HT Syndication.
