MUMBAI, India, Feb. 13 -- Intellectual Property India has published a patent application (202641007422 A) filed by SR University, Warangal, Telangana, on Jan. 26, for 'a transfer learning-based system for early detection of multiple crop diseases and pest infestations using leaf and field images.'
Inventor(s) include Dr. Sumit Jain; Dr. Rupesh Mishra; and Dr. Piyush Moghe.
The application for the patent was published on Feb. 13, under issue no. 07/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a transfer learning-based system (100) for predictive and preemptive detection of multiple types of crop diseases and insect pest infestations, applicable at both the plant level (leaf images (101)) and field level (field images (102)). The system integrates a modular framework comprising an image acquisition module (110), an image preprocessing module (120), a transfer learning-based deep learning model (130), a multi-class classification module (140), an infestation estimation unit (150), and an early warning and output module (160). The image acquisition module (110) is configured to capture diverse imagery of crops and agricultural environments using mobile devices, digital cameras, or aerial platforms. The captured images are processed through the image preprocessing module (120), which performs operations such as noise reduction, cropping, resizing, normalization, and enhancement, preparing the data for analysis by the deep learning model. The transfer learning-based deep learning model (130), built upon a pre-trained convolutional neural network and fine-tuned using agricultural image datasets (131), enables accurate detection of multiple diseases and pests even with limited labeled data. The multi-class classification module (140) identifies various crop diseases and pest infestations, while the infestation estimation unit (150) assesses the severity of the detected condition, categorizing it into early, moderate, or severe levels. The early warning and output module (160) generates timely alerts to facilitate informed crop health management and remedial actions. By combining transfer learning with multi-class classification and severity assessment, the invention provides an efficient, scalable, and real-time solution for early disease and pest detection in agriculture, reducing the dependency on extensive labeled datasets and supporting improved crop protection, yield, and sustainability."
Disclaimer: Curated by HT Syndication.
