Intellectual Property India Publishes Patent Application for Deep Learning-Based Industrial Visual Inspection System For Automated Defect Detection And Classification

MUMBAI, India, Oct. 5 -- Intellectual Property India has published a patent application (202641115611 A) filed by R. M. D. Engineering College on September 26, 2026, for Deep Learning-Based Industrial Visual Inspection System For Automated Defect Detection And Classification.

Inventors include Dr. A. Gnanasekar; Dr. J. Sherine Glory; Dr. A. Tamizharasi; Mr. T. Venkatesan; and Mrs. G. Manisha.

The application for the patent was published on October 02, 2026, under issue no. 40/2026.

Abstract: The present invention discloses a Deep Learning-Based Industrial Visual Inspection System for Automated Defect Detection and Classification comprising an image acquisition unit, controlled illumination unit, product positioning unit, image preprocessing module, deep learning processing module, defect detection module, defect classification module, decision-making module, industrial controller, rejection/segregation unit, data storage module, communication interface, and user interface. The image acquisition unit captures images of a product under controlled illumination, and the image preprocessing module performs normalization, enhancement, noise reduction, geometric correction, and region-of-interest extraction. The deep learning processing module analyzes the processed image to extract visual features and identify potential defect regions. The defect detection and classification modules determine the location and category of one or more defects and generate corresponding confidence values. The decision-making module evaluates the detection results using a predefined confidence criterion and generates an accept, reject, or secondary- inspection decision. The industrial controller can actuate a product segregation mechanism for automatically separating defective products from acceptable products. Inspection images, defect categories, locations, confidence values, timestamps, and product information may be stored for traceability and quality analysis. The invention thereby provides automated, consistent, real-time, and adaptable industrial visual inspection and defect classification.

Disclaimer: Curated by HT Syndication.