EXTRACTION OF ENERGY-INFLUENTIAL PARAMETERS FROM BUILDING FAÇADE IMAGES THROUGH GOOGLE STREET VIEW

Abstract

Energy modeling is a crucial tool at the city level for city managers to take decisions related to the building stock. To achieve this, urban models need additional building information to ensure good-quality simulations. Automated image analysis has shown potential in many fields but has lacked to appear in works aiming to improve urban energy analysis. Thus, the objective of this study is to provide a methodology for the extraction of the window-to-wall ratio from building façade images. The methodology proposed in this study includes training a semantic segmentation model. Results of this study have shown that image segmentation models have great potential in extracting the window-to-wall ratio from façade images.