Predicting the Solar Potential of Rooftops using Image Segmentation and Structured Data Daniel DE BARROS SOARES François ANDRIEUX Bastien HELL namR, Paris, France namR, Paris, France namR, Paris, France
[email protected] [email protected] [email protected] Julien LENHARDT Jordi BADOSA Sylvain GAVOILLE ENSTA Paris LMD, Ecole polytechnique, IP Paris namR, Paris, France namR, Paris, France Palaiseau, France
[email protected] [email protected] [email protected] Stéphane GAIFFAS Emmanuel BACRY LPSM, Université de Paris CEREMADE, Université Paris Dauphine DMA, Ecole normale supérieure namR, Paris, France namR, Paris, France
[email protected] [email protected] Abstract Estimating the amount of electricity that can be produced by rooftop photovoltaic systems is a time-consuming process that requires on-site measurements, a dif- ficult task to achieve on a large scale. In this paper, we present an approach to estimate the solar potential of rooftops based on their location and architectural characteristics, as well as the amount of solar radiation they receive annually. Our technique uses computer vision to achieve semantic segmentation of roof sections and roof objects on the one hand, and a machine learning model based on struc- tured building features to predict roof pitch on the other hand. We then com- pute the azimuth and maximum number of solar panels that can be installed on a rooftop with geometric approaches. Finally, we compute precise shading masks and combine them with solar irradiation data that enables us to estimate the yearly solar potential of a rooftop. arXiv:2106.15268v1 [cs.CV] 28 May 2021 1 Introduction The 21st century is characterized by an ever-increasing energy consumption and greenhouse gas emissions that are contributing to climate change in an unprecedented way.