Design Of A Remote Monitoring System For Smart Window Based On Cloud Platform

With the rapid development of the Internet of Things,smart products have gradually integrated into people’s lives.As an important part of the home,windows are constantly moving towards intelligence.Most of the existing smart window products are single-function and expensive,and have not yet entered commercialization.Through the analysis of existing technology and research,a remote monitoring system for smart windows based on cloud platform is proposed.STM32 is selected as the main control module in the system.Compared with traditional windows,a variety of control methods are designed in the system,including automatic control based on multi-sensors,remote control based on cloud platform,gesture control based on Field Programmable Gate Array(FPGA),button control and remote control.In the automatic control mode,the indoor and outdoor environment information is collected by multi-sensors in real time,which decide whether to open and close the window.Multi-sensors’ information is also sent to the Organic Light-Emitting Diode(OLED)display for real-time displaying.In the remote control mode,multi-sensors’ information and window status are uploaded to Alibaba Cloud platform for storage through the Wireless Fidelity(WiFi)module using the Message Queuing Telemetry Transport(MQTT)protocol.The interface and functions of the mobile phone and the web are designed,which can realize the remote monitoring of indoor and outdoor environment and window status,and can also publish control signals to the hardware to drive specific operation.In the gesture control mode,considering the limited memory of the main controller,FPGA is selected for gesture control.The core of gesture control is gesture recognition.First,the OV7670 camera is used to collect image information and transfer it to FPGA,and then go through the process of color space conversion,skin color detection,gesture positioning,gesture segmentation and gesture recognition.The system incorporates the lower edge positioning and segmentation method,and the jump edge detection method is used for gesture recognition for the first time.Finally,the recognized gesture signal is sent to the main controller through the Serial Peripheral Interface(SPI)protocol as a control signal for controlling.Gesture control also includes two parts,Synchronous Dynamic Random Access Memory(SDRAM)cache and Video Graphics Array(VGA)display,which can respectively improve the speed of image processing and the effect of real-time display.The test results show that the effects of multiple control methods in this system have reached expectations and meet the diverse needs of people.