Abstract: In the current environment of intelligent agriculture, the identification and monitoring of crop growth process has been a challenging task. Based on this, a remote greenhouse visual monitoring system based on Internet of things is proposed, the system monitors environmental parameters such as temperature, humidity and light intensity of crops in greenhouse by LoRa wireless communication technology, which can monitor the growth of crops in time, and to achieve the greenhouse crops automatic ventilation, automatic lighting and other functions. Real-time monitoring of environmental information and control of environmental parameters in the greenhouse from the Qt host on the PC, and the crops are monitored by OV9726 camera. The information obtained is transmitted to S3C6410 centralized control module for processing, a combination of clonal selection algorithms and Naive Bayes classifier was used to identify the leaves. This system uses LoRa module to implement environmental monitoring, and the Linux operating system is transplanted to the centralized control module to prepare for the construction of the hardware and software platform of the visual system, the combined algorithm can make the recognition rate of crop leaves reach 95.3% and the recognition time reach 8.4 ms, which improves the precision of leaf recognition obviously, the validity of the equipment and algorithm used in this system is fully verified by experiments.