MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202641073137 A) filed by Vellore Institute Of Technology, Chennai on June 12, 2026, for Real-Time Citrus Fruit Disease Classification System Using Feature Recalibration Attention-Based Deep Learning.
Inventors include Sankar P; Sahil Sandeep; and Shivanya Shomir Dutta.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: The present disclosure relates to a system (102) and method (400) for real-time classification of citrus fruit diseases. The system (102) includes a control unit (108) can include a processor (202) and a memory (204). The control unit (108) receives image data of a citrus fruit from an image acquisition unit (104). The received image data is preprocessed to produce pre-processed image data. Disease-specific features are extracted from the pre-processed image data through a feature recalibration neural processing model stored in the memory (204), wherein channel-wise reweighting amplifies disease-relevant features. The extracted disease-specific features are classified into predefined disease categories to determine the disease condition of the citrus fruit. A disease classification result and a confidence score are generated through an output unit (220). All processing operations are executed on the processor (202) as a self-contained on-device classification operation.
Disclaimer: Curated by HT Syndication.
