Laboratory Monitoring System Based On Linux

As the main place for teaching practice and scientific research in higher education institutions,laboratories are densely populated and highly mobile,and they also hold many precision instruments and store various flammable and explosive dangerous items,so safety accidents occur from time to time,which seriously threaten the personal safety of teachers and students and cause huge losses to school property.Therefore,the safety monitoring of laboratories has become the focus of increasing attention.Traditional laboratories adopt the safety monitoring mode of manual guarding,which is costly and inefficient.In this context,in order to break the traditional bondage and improve the intelligent monitoring of laboratories,this article proposes and designs a laboratory monitoring system based on Linux.The system as a whole is based on embedded technology,using Radxa ROCK 5B as an embedded terminal,by transplanting Ubuntu 20.04 operating system,combining various hardware devices and various technical means to achieve the design requirements,and finally achieving the purpose of laboratory safety monitoring.The system uses ZigBee technology to form a wireless network inside the lab.The coordinator node establishes communication with the embedded terminal device through serial port to USB to upload the collected data and send the control commands.The terminal node is equipped with DHT11 temperature and humidity sensor,MQ-2gas sensor and flame sensor to collect data from the site environment,and cooperate with various linkage devices to complete the task of environmental monitoring.After repeated tests,it is found that once the environment changes abnormally,it can sense and trigger the device response in time,with high detection accuracy and low time delay.At the same time,the system combines RFID technology and 4G technology to achieve access control and SMS phone alarm function to ensure efficient operation of the laboratory safety monitoring system.On this basis,the system introduces target detection technology for effective identification of laboratory smoking,a violation of the use of fire.Based on the YMIR open source algorithm platform,the smoking detection network model is trained using the YOLOv5 s algorithm,and the lightweight network model obtained from the training is deployed in the embedded terminal to achieve rapid identification of smoking behavior in the laboratory.The trained deep learning network model has a good recognition effect with an accuracy rate of 95.06%,and the processing time of each smoking picture is only about 60 ms after it is deployed in the embedded terminal.The lightweight recognition system not only ensures high recognition accuracy,but also achieves rapid recognition of smoking behavior,which can effectively assist in laboratory safety monitoring.Finally,the system uses Qt technology to design a friendly human-computer interaction interface,which realizes the interface functions such as laboratory access control,SMS phone alarm,environmental data display,linkage equipment control,and real-time detection of laboratory smoking behavior,providing a visual window for users and further enhancing the level of laboratory intelligent monitoring.