The multi-modal two-step floating catchment area (MM-2SFCA) method is an extension of the two-step floating catchment area (2SFCA) method that incorporates the impact of different transportation modes, thereby facilitating more accurate calculations of the spatial accessibility of public facilities in urban areas. However, the MM-2SFCA method does not account for the impact of distance within the search radius on supply–demand capacities, and it assumes an idealized supply–demand relationship. This paper introduces the gravity model into the MM-2SFCA method, proposing a multi-modal gravity-based 2SFCA (MM-G2SFCA) method to better account for distance decay and supply–demand relationships. Furthermore, a standardized gravity model is proposed based on the traditional gravity model. This model imposes constraints on upper and lower limits for distance decay weights without compromising the fundamental curve characteristics of the gravity model, thereby avoiding extreme weight scenarios. The accessibility of public hospitals in Shenzhen is evaluated through the integration of basic geographic information data, resident travel data, and official statistical data. The findings demonstrate that the standardized gravity model effectively addresses the issue of excessively high local distance weights in the traditional gravity model, making it more suitable as a distance decay function. The MM-G2SFCA method improves the consideration of distance and supply–demand relationships, thereby facilitating a more rational distribution of accessibility on a global scale. This study discovers differences in the spatial allocation of public hospital resources across the Shenzhen’s districts. Accessibility within the metropolitan core is significantly higher than that outside the core. Additionally, there is a notable difference in the level of accessibility among the districts. Accessibility is found to be better in district centers and along the main traffic arteries.

Recently, unsupervised domain adaptation (UDA) segmentation of remote sensing images (RSIs) has attracted a lot of attention. However, the performance of such methods still lags far behind that of their supervised counterparts. To this end, this paper focuses on a more practical yet under-investigated problem, semi-supervised domain adaptation (SSDA) segmentation of RSIs, to effectively improve the segmentation results of targeted RSIs with a few labeled samples. First, differently from the existing single-prototype mode, a novel cross-domain multi-prototype constraint is proposed, to deal with large inter-domain discrepancies and intra-domain variations. Specifically, each class is represented as a set of prototypes, so that multiple sets of prototypes corresponding to different classes can better model complex inter-class differences, while different prototypes within the same class can better describe the rich intra-class relations. Meanwhile, the multi-prototypes are calculated and updated jointly using source and target samples, which can effectively promote the utilization and fusion of the feature information in different domains. Second, a contradictory structure learning mechanism is designed to further improve the domain alignment, with an enveloping form. Third, self-supervised learning is adopted, to increase the number of target samples involved in prototype updating and domain adaptation training. Extensive experiments verified the effectiveness of the proposed method for two aspects: (1) Compared with the existing SSDA methods, the proposed method could effectively improve the segmentation performance by at least 7.38%, 4.80%, and 2.33% on the Vaihingen, Potsdam, and Urban datasets, respectively; (2) with only five labeled target samples available, the proposed method could significantly narrow the gap with its supervised counterparts, which was reduced to at least 4.04%, 6.04%, and 2.41% for the three RSIs.

Residential areas and road networks have a strong geographical correlation. The development of a single geographical feature could destroy the geographical correlation. It is necessary to establish collaborative generalization models suitable for multiple features. However, existing road network generalization methods for mapping purposes do not fully consider residential areas. Compared with road networks, residential areas have a higher priority in cartographic generalization. In this regard, this study proposes a road network generalization method constrained by residential areas. First, the roads and settlements obtained from clustering residential areas were classified. Next, the importance of the settlements was evaluated and certain settlements were selected as the control features. Subsequently, a geographical network with the settlements as the nodes was built, and the traffic paths between adjacent settlements were searched. Finally, redundant paths between the settlements were simplified, and the visual continuity and topological connectivity were checked. The data of a 1:50,000 road network and residential areas were used as the experimental data. The experimental results demonstrated that the proposed method preserves the overall structure and relative density characteristics of the road network, as well as the geographical correlation between the road network and residential areas.

The division of urban functional zones is crucial for understanding urban characteristics and aiding in urban management and planning. Traditional methods, like dividing based on blocks and grids, are insufficient for modern demands. To address this, a knowledge-graph-supported method for building functional category division is proposed. Firstly, the associations between points of interest (POI) and buildings are established using triangulation and buffer zones. Then, a knowledge graph of buildings is constructed through entity and relationship extraction. A functional category classification model supported by the Z-score is designed using the semantic characterizations of surrounding POIs for inference rules. The results demonstrate high accuracy in building functional category division, supporting the refinement and intelligent expression of urban functional zones for urban construction, planning, and management.

This study focuses on a skeleton line extraction algorithm of areal hydrographic elements for cartographic generalization. The requirements of skeleton line extraction for areal hydrographic elements were first analysed and a new automatic extraction algorithm that considered spatial and hierarchical relationships was proposed to meet these requirements. A graph with structured areal hydrographic elements was subsequently presented, and the graph pruning operation was conducted based on the spatial relationship and pruning length threshold. The skeleton line connection of the traditional unconstrained edge midpoint algorithm was improved by the guidance of spatial and hierarchical relationships. Finally, the proposed method was used to extract skeleton lines from 1:250,000 hydrographic elements, and compared with other algorithms. The experiments demonstrate that the skeleton lines extracted by the proposed method have certain advantages in natural continuity, hierarchy, and topological consistency; the results are closer to those of manual extraction and avoiding the subsequent optimization process.