2025 IFAL Invitation | 1.65MB | Download |
Prediction model for indoor light environment brightness based on image metricsPrediction model for indoor light environment brightness based on image metrics Chao Ruana , Li Zhoua,b , Liangzhuang Weia , Wei Xub , Yandan Lina,b,c,* aAcademy for Engineering and Technology, Fudan University, Shanghai 200433, China b lnstitute for Electric Light Sources, School of Information Science and Technology, Fudan University, Shanghai 200433, China cInstitute for Six-sector Economy, Fudan University, Shanghai 200433, China
Abstract Rapid progress in display technology and optical simulation software has enabled the visualization of lighting design, which can provide abundant visual information. However, renderings only allow designers to subjectively judge whether the lighting layout and optical parameters are reasonable. So we want to combine the rendered images and photometric data in the process of optical simulations to define an evaluation indicator of spatial brightness, which can quantify the perceived brightness of the simulated scene. An image assessment experiment based on a display was conducted to investigate the relationship between spatial brightness and calculated image metrics of indoor lit environments. Participants evaluated spatial brightness perception of 39 images of indoor lit environments simulated with SPEOS simulation software on the screen. Three metrics (Average luminance, RAMMG contrast, correlated color temperature (CCT)) were used to characterize participants’ spatial brightness scores, and the relevant prediction equations were proposed. The application of the RAMMG contrast to spatial brightness prediction has a good performance. The image-based assessment method developed in this study has a high Pearson’s correlation coefficient (rp=0.935) with the actual visual assessment, which is reliable and convenient. The proposed model performs better compared with other prediction methods available. |