Building Mapping

Building mapping focuses on automatically generating accurate 2D and 3D representations of buildings from various data sources, primarily high-resolution satellite and aerial imagery, and LiDAR data. Current research emphasizes the use of deep learning, particularly convolutional neural networks and transformers, to improve the accuracy and efficiency of building extraction, segmentation, and polygonization, often incorporating multi-task learning to leverage correlations between buildings and other urban features like roads. These advancements are crucial for applications ranging from urban planning and disaster response to climate modeling and autonomous navigation, enabling more precise and comprehensive analyses of built environments.