Modelling Lighting and Spatial Characteristics of Indoor Environments for Adaptive Building Systems
Restricted (Penn State Only)
- Author:
- Wang, Yuwei
- Graduate Program:
- Architectural Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 02, 2024
- Committee Members:
- James Freihaut, Program Head/Chair
Dorukalp Durmus, Chair & Dissertation Advisor
Hui Yang, Outside Unit & Field Member
Julian Wang, Major Field Member
Yuqing Hu, Major Field Member - Keywords:
- visual perception;
indoor lighting;
image quality metrics;
intelligent building systems;
autonomous lighting; - Abstract:
- Estimating perceived visual quality has significant benefits, enabling efficient lighting systems that reduce energy use, boost productivity, and enhance indoor well-being through dynamic adjustments based on environmental sensing to optimize visual perceptions. However, implementing such intelligent systems requires initially estimating human responses to ensure lighting adjustments align with user preferences and needs. For intelligent building systems to attain widespread success, a more profound comprehension of architectural environments and human reactions is imperative. Examining the spatial characteristics of architectural environments, including visual scenes and objects, can provide valuable insights into human visual perceptions. This aspect, often overlooked in prior studies, is crucial for accurate analysis. In this dissertation, three perceptual quality assessment experiments were conducted to evaluate the spatial characteristics, lighting, and color quality to predict occupants’ responses to the indoor environment. In the first experiment, 60 images of different spatial spaces were exhibited on a display to assess the accuracy of image quality metrics in estimating the visual complexity, visual clarity, visual preference, and colorfulness for images of indoor environments. The second experiment evaluated the effects of lighting (background horizontal illuminance, display, and background luminance) on subjective evaluations of 12 indoor environment images that were shown on a display under 8 different lighting conditions. New metrics were developed to predict visual quality based on lighting and spatial variations. Finally, the third experiment assessed the performance of new metrics to predict perceived visual quality in non-virtual immersive indoor environments. A set of different metrics was developed to predict the perceived visual quality of immersive indoor environments based on photometric, colorimetric, and image quality metrics. Overall, the dissertation investigates the effectiveness of image quality metrics to evaluate the perceived quality of indoor environments and their potential applications in adaptive lighting systems. Future research should explore the robustness of the computational metrics across a broader set of indoor settings, lighting conditions (including spatial and spectral light distributions), and user cognitive and psychological responses. Investigating how subjective human evaluations and objective metrics can be integrated, along with exploring human factors such as age, visual perception, cognitive abilities, and cultural backgrounds, could provide valuable insights for developing comprehensive, user-centric intelligent lighting systems.