Liangsheng Shi

School of Water Resources and Hydropower Engineering, Wuhan University

It is difficult to standardize production management in a changing environment for agriculture. JiaHe system is an artificial intelligence system for agricultural water and fertilizer management. It uses smart phones as the main carrier, uniting sensors, smart phone, UAV and satellite, and less dependent on expensive sensor investment. Machine vision technique for diagnosing crop is developed by mining texture and spectral traits of phone images and utilizing environmental information. A new Ground-Air-Space (GAS) perception neural network is embedded in the system for lost-cost and high-accuracy perception of regional crop growth status. We also propose a hybrid data assimilation architecture which can simultaneously assimilate sensor data, phone and UAV images. JiaHe system has been successfully applied to paddy rice planting management in Southern China.