Research On Intelligent Image Sensing Technology In Lighting Control System

The intelligentization of lighting control technology plays an important role in energy conservation and emission reduction.However,the infrared sensing and sound sensing technologies commonly used in lighting control systems have low recognition accuracy for the activities of people in the lighting scene,and the detection distance of the sensors is close.With the development of machine vision technology and deep learning theory,image recognition technology is expected to make up for the above shortcomings in the lighting control system.But the current deep learning model has too high requirements on data volume,computing power and training time,which has caused certain difficulties for the development of image sensing systems.Aiming at the above problems,an intelligent image sensing based on image recognition is designed.This image sensing mainly recognizes the activity of the person in the image in the illumination scene,and controls the illumination system by the recognition result.In the realization process of intelligent image sensing technology,the deep learning process is optimized from the aspects of image processing and neural network model,which overcame the shortcomings of the current deep learning model in terms of data volume,computational power and training cycle,thereby enabling image sensing to achieve faster response speed.In the aspect of image processing,a dynamic template matching strategy is proposed,which improves the pertinence and robustness in the image recognition process.In the construction of neural network model,the structure and parameters of the neural network are optimized,which reduces the parameter quantity and calculation amount of the model training.Finally,the image sensor was transplanted into embedded processor,the recognition accuracy and response speed of the system are tested.And combined with the specific lighting control environment,the use of image sensing has been verified.The research results show that intelligent image sensing can accurately and quickly detect the signs of human activities in the lighting environment.Compared with traditional sensing devices,image sensing can cover a large range and realize distributed monitoring of different illumination areas.