The traditional paradigm for studying urban morphology involves the interpretation of Nolli maps, using methods such as morphometrics and visual neural networks. Previous studies on urban morphology discovery have always been based on raster analysis and have been limited to the central city area. Raster analysis can lead to fragmented forms, and focusing only on the central city area ignores many representative urban forms in the suburbs and towns. In this study, a vast and complex dataset was applied to the urban morphology discovery based on the administrative community or village boundary, and a new image deformation pipeline was proposed to enhance the morphological characteristics of building groups. This allows visual neural networks to focus on extracting the morphological characteristics of building groups. Additionally, the research on urban morphology often uses unsupervised learning, which means that the learning process is difficult to control. Therefore, we refined the NT-Xent loss so that it can integrate morphological indicators. This improvement allows the visual neural network to “recognize” the similarity of samples during optimization. By defining the similarity, we can guide the network to bring samples closer or move them farther apart based on certain morphological indicators. Three Chinese cities were used for our testing. Representative urban types were identified, particularly some types located at the urban fringe. The data analysis demonstrated the effectiveness of our image deformation pipeline and loss function, and the sociological analysis illustrated the unique urban functions of these urban types.

The photorealistic reconstruction and rendering of architectural scenes have extensive applications in industries such as film, games, and transportation. It also plays an important role in urban planning, architectural design, and the city's promotion, especially in protecting historical and cultural relics. The 3D Gaussian Splatting, due to better performance over NeRF, has become a mainstream technology in 3D reconstruction. Its only input is a set of images but it relies heavily on geometric parameters computed by the SfM process. At the same time, there is an existing abundance of raw 3D models, that could inform the structural perception of certain buildings but cannot be applied. In this paper, we propose a straightforward method to harness these raw 3D models to guide 3D Gaussians in capturing the basic shape of the building and improve the visual quality of textures and details when photos are captured non-systematically. This exploration opens up new possibilities for improving the effectiveness of 3D reconstruction techniques in the field of architectural design.

Urban modeling is essential for city planning, scene synthesis, and gaming. Existing image-based methods generate diverse layouts but often lack geometric continuity and scalability, while graph-based methods capture structural relations yet overlook parcel semantics. We present a controllable framework for large-scale 3D vector urban layout generation, conditioned on both geometry and semantics. By fusing geometric and semantic attributes, introducing edge weights, and embedding building height in the graph, our method extends 2D layouts to realistic 3D structures. It also enables users to directly control the output by modifying semantic attributes. Experiments show that it produces valid, large-scale urban models, offering an effective tool for data-driven planning and design.