This deep learning model is designed to extract building footprints from high-resolution imagery (15–25 cm). Building footprint layers are essential for creating basemaps and supporting analysis workflows in urban planning, infrastructure development, insurance, taxation, change detection, and more.

Traditionally, building footprints are manually digitized from imagery—a process that is both time-consuming and labour-intensive. Deep learning offers a powerful solution by learning complex spatial patterns and delivering highly accurate results. By automating the extraction process, this model significantly reduces the time and effort needed to generate precise building footprint data.