MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641104365 A) filed by Srm University- Ap on August 31, 2026, for System And Method For Compound Image Degradation Classification.
Inventors include Dr. Rajiv Senapati; and Mohanty, Stutee.
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: A system (100) for compound image degradation classification includes a pre-processing module (102) configured to receive an input image and perform resize and normalization operations to generate a pre-processed image. A shared convolutional residual feature extractor (104) is connected to the pre-processing module (102) and extracts feature representations through sequential convolutional blocks. A global average pooling layer (106) aggregates spatial information into a pooled feature representation. A classification head (108) implemented as a fully connected neural network outputs multi-label classification predictions identifying degradation categories including motion blur, defocus blur, Gaussian noise, salt and pepper noise, JPEG artifacts, and shadow. A regression head (110) implemented as a fully connected neural network outputs a continuous severity level value indicating degradation magnitude. The classification head (108) and regression head (110) share the pooled feature representation, enabling simultaneous identification of multiple co-existing degradation types and estimation of continuous severity levels within a unified architecture.
Disclaimer: Curated by HT Syndication.
