Research and design of classroom lighting system based on deep learning
Aiming at the poor sensitivity of classroom lighting control method by the traditional sensor,as well as the misdetection and omission problem of personnel detection algorithm in the classroom,in order to fully and reasonably use the classroom lighting equipment resource,a classroom lighting system based on improved YOLOv5 and regional positioning is proposed.The system judges the sub-region where classroom personnel are located and generates control information for lighting equipment.Through training and testing on the self-made data set,and the build LED dot matrix simulates the lighting equipment in the classroom.The results show that the accuracy of YOLOv5 reaches 93%and the recall rate reaches 96%.It can accurately control the lighting equipment in the classroom in real time.The system has convenience to lighting,and saves energy.
deep learningYOLOv5classroom personnel detectregional positioning
万方数据
维普
