MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641118200 A) filed by Sathyabama Institute Of Science And Technology on October 02, 2026, for Planthealth-Vit: A Cnn–transformer Framework For Intelligent Plant Disease Detection And Crop Health Assessment.
Inventor includes Dr. M. Emimal.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: PlantHealth-ViT: A CNN–Transformer Framework for Intelligent Plant Disease Detection and Crop Health Assessment presents an intelligent deep-learning framework for automated plant disease detection and comprehensive crop health assessment. Conventional plant disease identification methods often depend on manual inspection, expert knowledge, and visible symptoms, which can be time-consuming and difficult to scale across large agricultural fields. To address these limitations, the proposed framework integrates the complementary capabilities of Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) to learn both fine-grained visual characteristics and global contextual relationships from plant images. Initially, crop images are subjected to preprocessing operations such as resizing, normalization, noise reduction, and data augmentation to improve image quality and model robustness. The CNN component extracts local features including leaf texture, color variations, edges, spots, lesions, and other disease-specific patterns, while the Vision Transformer captures long-range spatial dependencies and broader relationships among different regions of the plant image. The extracted representations are subsequently combined through a feature-fusion mechanism to generate a comprehensive plant-health representation. The fused features are utilized for automated disease classification and crop health assessment, enabling the identification of healthy and diseased plants and classification of different disease categories. An explainable AI component can further be incorporated to generate visual explanations of the model's predictions, helping users identify the regions of the plant that contributed to the classification decision. The proposed framework is intended to provide an efficient, scalable, and interpretable solution for intelligent crop monitoring and precision agriculture applications.
Disclaimer: Curated by HT Syndication.
