Abstract
Glare is a common local visual discomfort that is difficult to identify with conventional light sensors. This article presents an artificial intelligence algorithm that detects subjective local glare discomfort from the image analysis of the video footage of an office occupant’s face. The occupant’s face is directly used as a visual comfort sensor. Results show that it can recognize glare discomfort with around 90% accuracy. This algorithm can thus be at the basis of an efficient feedback control system to regulate shading devices in an office building.
Highlights
Enhancing the indoor daylight conditions is a key issue to address as it greatly impacts building occupants comfort and satisfaction, in a working environment
12 test participants are used for training and testing the prototype of the glare detection algorithm. 2 different strategies have been investigated to split the experimental dataset into training and test data
A glare detection algorithm has been developed with off-the-shelf pre-trained neural networks that can readily be used for various image analysis and object detection
Summary
The aim of the project presented in this article is to develop a proof of concept prototype to demonstrate that the subjective visual comfort of an office occupant can be accurately assessed by a computer algorithm analysing the video footage of the human’s face. The output of such an algorithm can be used as feedback signal to control dimmable artificial lighting and motorized solar shading devices to optimize the effective local visual comfort of the occupants. This paper presents the different steps for the development of a glare detection algorithm based on facial image analysis.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.