The intelligent greenhouse environment monitoring system designed and implemented in this study utilizes the Waffle Nano development board, runs on the OpenHarmony operating system, and employs the Message Queuing Telemetry Transport protocol as the communication protocol. The EMQX Cloud server serves as a bridge to communicate with the remote monitoring APP for the greenhouse environment designed in this system. Additionally, a CNN-LSTM hybrid neural network is used to specifically analyze the correlation between greenhouse environment data and specific variables, as well as the impact of occupancy status on the distribution of variables. The system performs functions such as greenhouse environment data monitoring, environmental equipment control, and remote monitoring. Testing showed that the system attained a temperature monitoring accuracy of 98.8%, which is an improvement of 3.6% over traditional temperature monitoring instruments. For humidity monitoring, the accuracy was 98.6%, an improvement of 1.9% over traditional humidity monitoring instruments, with the equipment control delay time being within 5 s. This system has significantly enhanced the accuracy and real-time response speed of greenhouse environment monitoring, and promoted the automation level of greenhouse management.