MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641095645 A) filed by Mohan Babu University on August 07, 2026, for Automatic Road Damage Extraction And Alert Generation.
Inventors include Ms. Alla Meghana Reddy; Ms. Tirumala Muni Kusumitha; Ms. Nadella Chandini; Ms. Yerrannagari Bhavigna; and Dr. M Sowmya Vani.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The road infrastructure monitoring system uses its two main functions to enhance traffic safety while it reduces expenses for road repairs. The project develops an automated system that uses deep learning technology to identify road damage through the combined application of YOLOv9 and Detection Transformers (DETR) algorithms. The Roboflow Road Damage Detection dataset includes seven types of road damage, which the model uses to identify alligator cracking, edge cracking, longitudinal cracking, transverse cracking, patching, potholes, and rutting. The system uses high-resolution road images to perform real-time surface defect analysis because YOLOv9 enables object detection while DETR-based models extract features and perform classification. The researchers can study the road extraction process because it creates road sections that they can study with complete focus. Damage assessment establishes permanent maintenance recommendations through its ability to assess both the damage extent and the damage type. The system shows major progress in inspection speed compared to traditional manual inspection methods. Django uses Google Colab as its training and inference platform, while Python operations run on the platform and results display through its front end. The system uses an automated alert mechanism that sends email alerts to the administrator when it detects vital damages because this allows for immediate action and scheduled maintenance work.
Disclaimer: Curated by HT Syndication.
