Real-time monitoring and assessment of indoor illuminance are crucial for maintaining environmental quality, visual comfort, and building energy efficiency. Traditional systems struggle with complex layouts, but integrating real-time monitoring with smart building management enhances comfort and sustainability. This study presents an innovative method combining a CMOS camera, data processing, and deep learning algorithms. The camera captures continuous images and extracts texture features affected by lighting. A machine learning model then predicts the illuminance distribution, generating detailed light maps in real-time. This novel sensing system, validated through experiments at the University of Skikda, provides a promising solution to improve light management by optimizing daylight use, reducing energy consumption, and enhancing visual comfort. It offers architects and facility managers a tool to integrate real-time data into building management systems for more sustainable indoor environments.
Read full abstract