MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641108190 A) filed by S. Hrushikesava Raju; Arpit Jain; T. Aswini Devi; and Koneru Lakshmaiah Education Foundation on September 09, 2026, for Cottonseveritynet: An Ensemble Model For Cotton Leaf Disease Severity Prediction.
Inventors include Arpit Jain; and T. Aswini Devi.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: The production of cotton leaf crops has declined significantly in recent years, affecting fiber quality due to severe cotton diseases. Hence, prevent loss and increase profitability by identifying the disease at an early stage. To predict, many existing methods rely on handcrafted features such as histograms, texture, shape, color, and LBP and GLCM attributes. These features fail in real-time scenarios due to background noise, occlusions on the leaf, irregular illumination, overlapping disease stages, and shadows. Methods such as Random Forest, SVM, CNN-based, and ANFIS face issues including the complexity of fuzzy rules, limited generalization, reduced performance, and weak feature representation when predicting multi-class diseases. Hence, a hybrid ensemble, namely CottonSeverityNet, is proposed, comprising efficient feature extraction for accurate discrimination, EfficientNetV2 for local lesion pattern extraction, ConvNeXT with global context attention for extracting global disease-spread patterns on the l
Disclaimer: Curated by HT Syndication.
