MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641107315 A) filed by Kkr & Ksr Institute Of Technology And Sciences on September 07, 2026, for Wad-Yolo: Wavelet-Based Adaptive Defect Detection With Multi-Resolution Feature Analysis For Steel Surfaces.
Inventors include Mr. Prof. Repudi Ramesh; Mrs. Dr. S. Sri Lakshmi Parvathi; A. Nagaraju; B. Chandra Sekhar; Ch. Moses Christopher; and B. Kotaiah.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: ABSTRACT [0011] A wavelet-based adaptive defect detection framework designated WAD-YOLO is disclosed for accurate identification of steel surface defects. Discrete wavelet transform decomposes industrial surface images into multiple frequency bands, thereby enhancing defect-related texture features at different spatial scales. The resulting multi-resolution descriptors are fused with the feature maps of a YOLO object-detection network, while recursive feature elimination retains only the most discriminative texture components. The integrated architecture improves detection of small, low-contrast defects such as cracks, scratches, pits and inclusions while preserving real-time processing speed. Experimental evaluation demonstrates superior performance relative to conventional convolutional detectors, rendering the system suitable for continuous industrial quality-control deployment.
Disclaimer: Curated by HT Syndication.
