<p><strong>Abstract.</strong> The paper describes a workflow for generating LoD3 CityGML models (i.e. semantic building models with structured facades) based on textured LoD2 CityGML models by adding window and door objects. For each wall texture, bounding boxes of windows and doors are detected using “Faster R-CNN”, a deep neural network. We evaluate results for textures with different resolutions on the ICG Graz50 facade dataset. In general, detected bounding boxes match very well with the rectangular shape of most wall openings. Thus, no further classification of shapes is required. Windows are typically aligned to rows and columns, and only a few different types of windows exist for each facade. However, the neural network proposes rectangles of varying sizes, which are not always aligned perfectly. Thus, we use post-processing to obtain a more realistic appearance of facades. Window and door rectangles get aligned by solving a mixed integer linear optimization problem, which automatically leads to a clustering of these openings into few different classes of window and door types. Furthermore, an a-priori knowledge about the number of clusters is not required.</p>
Abstract. A challenge in data-based 3D building reconstruction is to find the exact edges of roof facet polygons. Although these edges are visible in orthoimages, convolution-based edge detectors also find many other edges due to shadows and textures. In this feasibility study, we apply machine learning to solve this problem. Recently, neural networks have been introduced that not only detect edges in images, but also assemble the edges into a graph. When applied to roof reconstruction, the vertices of the dual graph represent the roof facets. In this study, we apply the Point-Pair Graph Network (PPGNet) to orthoimages of buildings and evaluate the quality of the detected edge graphs. The initial results are promising, and adjusting the training parameters further improved the results. However, in some cases, additional work, such as post-processing, is required to reliably find all vertices.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.