MUMBAI, India, Feb. 6 -- Intellectual Property India has published a patent application (202641007352 A) filed by Keshav Memorial Engineering College, Hyderabad, Telangana, on Jan. 25, for 'deepagrinet: cnn-lstm-based crop yield forecasting using multisource agricultural data.'
Inventor(s) include Dr. Pillareddy Vamsheedhar Reddy; Mrs. Ratcha Jamuna; Mrs. Kota Harini; Mrs. Gayathri Tippani; Mr. M Anil Kumar; Mr. V Maddileti Reddy; Dr. Sunil Kumar Thota; and Dr. Ch Rathan Kumar.
The application for the patent was published on Feb. 6, under issue no. 06/2026.
According to the abstract released by the Intellectual Property India: "Proper forecasting of crop yields is essential to food security, agricultural planning, and stability in supply-chain; nonetheless, the conventional methods of yield forecasting have remained low yield forecasting failures due to dynamic climatic and environmental factors. The traditional statistics approach and single source machine learning models are usually based on historical averages or single parameter which only gives low levels of prediction (70-80%) and low levels of early stage forecasting (55-60%) mostly when there was an extreme weather and when weather changes. In addition, the models do not collectively model spatial crop health patterns and long-term temporal dependencies of multisource agricultural data. In order to overcome the stated limitations, the current work suggests Deep Agricultural Intelligence Network (DeepAgriNet) as an integrated CNN-LSTM-based crop yield prediction application that incorporates satellite imagery, weather time-series data, as well as soil characteristics and historical yield data in a single deep learning model. Massive testing of various crops (rice, wheat, maize, and soybean) in ten growing seasons shows that DeepAgriNet can predict the yields at an average of 96.8, early-stage predictability of 91, and much lower MAE and RMSE than standard machine learning and statistical models. The suggested system is also highly scalable, automated, and resilient to the unpredictable climatic conditions. In general, DeepAgriNet is a promising, climate-responsive, and data-driven approach to precision agriculture, which allows to estimate early yields with high accuracy and optimize resources, as well as make sustainable decisions in agriculture."
Disclaimer: Curated by HT Syndication.
