MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202611054449 A) filed by Meerut Institute Of Engineering & Technology, Meerut; Ms. Aarti Verma; Dr. Punit Mittal; Neeti Bhardwaj; Vanshika Raheja; Yuvraj Sharma; and Vanshika Rana on April 29, 2026, for A Real-Time Multi-Product Industrial Defect Detection And Classification System Using Integrated Deep Learning Pipeline.
Inventors include Ms. Aarti Verma; Dr. Punit Mittal; Neeti Bhardwaj; Vanshika Raheja; Yuvraj Sharma; and Vanshika Rana.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: The present invention relates to an automated industrial inspection system for real-time detection and classification of defects in multiple product categories using computer vision and deep learning techniques. The system processes video input from production lines, extracts frames, and performs preprocessing including resizing, normalization, and binary mask generation. A real-time object detection module based on YOLO architecture identifies and localizes multiple objects within each frame. Detected objects are then isolated and forwarded to a convolutional neural network-based classification module for defect determination. The system integrates image processing, feature extraction, and classification into a unified pipeline, enabling simultaneous localization and defect identification. A visualization module overlays bounding boxes with defect indicators on output frames. The invention supports multiple product types including cashews, capsules, candles, and fryums, and achieves consistent high accuracy while reducing reliance on manual inspection in industrial environments.
Disclaimer: Curated by HT Syndication.
