Research on Classroom Lighting Automatic Control System Based on Personnel Target Detection

Abstract

. With the popularity of digital campus and the proposal of building a conservation campus, classroom lighting power saving is an important expense of electricity consumption in universities. Since most classrooms are equipped with cameras, this paper proposes an automatic classroom lighting control system based on personnel target detection to achieve intelligent lighting control and power saving. The main research consists of three parts. One is to study and optimize the YOLO personnel target detection algorithm, and use the data set for model training to improve the recognition accuracy of the system. The second is to build the hardware platform of the system, using the Raspberry Pi with GPU as the controller to improve the image processing speed. The third is to carry out the control flow design of the light groups and the design of the image display window to realize the centralized seating and lighting control of classroom. Classroom personnel identification experiments and ambient light brightness detection experiments were carried out respectively, and the experimental results showed that the system had faster response time consumption and higher classification accuracy. The system is simple and cost-saving to implement with the help of existing equipment, and the study has certain practical application value and universality.