MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641079718 A) filed by Sona College Of Technology on June 29, 2026, for A System And Method For Non-Destructive Visual Estimation Of Gold Purity Using Deep Neural Networks.
Inventors include Kaladevi A C; Lathika V; Linkeshwara N S; Loga Priya M; Kanishka Devi R; and Lashwarth K.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: ABSTRACT The present invention discloses a computer-implemented system and method for non-destructive estimation of gold authenticity and purity using artificial intelligence and computer vision techniques. The system receives one or more digital images of a jewelry article through a camera-enabled device and performs image preprocessing including resizing, normalization, and noise reduction. A deep neural network-based classification model extracts visual features and determines whether the article is gold or non- gold. Upon positive identification, a visual purity estimation module analyzes the extracted feature vectors using clustering-based machine learning techniques to estimate a corresponding purity category. The system further generates confidence scores and risk indicators based on classification probability and cluster proximity metrics. The final output includes material type, estimated karat level, confidence percentage, and risk assessment. The invention provides a rapid, portable, contactless, and cost-effective alternative to conventional gold testing methods without requiring chemical testing, destructive sampling, or specialized laboratory equipment.
Disclaimer: Curated by HT Syndication.
