Greenhouse management system based on multi-node data fusion and visual recognition

With the rapid development of smart agriculture, especially in managing greenhouse greenhouses with large areas, the current popular monitoring methods have limitations in their measurements. To make the system monitoring more comprehensive and remote control more convenient, this paper proposes a more integrated system design. This monitoring system uses ZigBee wireless communication technology for data transmission, multi-node data processing through data fusion algorithms, training YOLOv5 model for plant leaf pest and disease identification, uploading data to the IoT cloud platform for storage, and remotely controlling peripherals such as greenhouse lights, windows, and so on through APP. Based on the tomato planting environment for system debugging, monitoring the tomato growing environment and leaf diseases, disease recognition accuracy is about 90%, and farmers can adjust the planting mode in time, to realize science and technology to help farmers.