Road extraction from high-resolution remote sensing images is a critical task in image understanding and analysis, yet it poses significant challenges because of road occlusions caused by vegetation, buildings, and shadows. Deep convolutional neural networks have emerged as the leading approach for road extraction because of their exceptional feature representation capabilities. However, existing methods often yield incomplete and disjointed road extraction results. To address this issue, we propose CR-HR-RoadNet, a novel high-resolution road extraction network that incorporates local and global context reasoning. In this work, we introduce a road-adapted high-resolution network as the feature encoder, effectively preserving intricate details of narrow roads and spatial information. To capture multi-scale local context information and model the interplay between roads and background environments, we integrate multi-scale features with residual learning in a specialized multi-scale feature representation module. Moreover, to enable efficient long-range dependencies between different dimensions and reason the correlation between various road segments, we employ a lightweight coordinate attention module as a global context-aware algorithm. Extensive quantitative and qualitative experiments on three datasets demonstrate that CR-HR-RoadNet achieves superior extraction accuracy across various road datasets, delivering road extraction results with enhanced completeness and continuity. The proposed method holds promise for advancing road extraction in challenging remote sensing scenarios and contributes to the broader field of deep-learning-based image analysis for geospatial applications.

The urban heat island (UHI) effect, a significant environmental challenge within the global urbanization process, poses severe threats to human health, ecological security, and life safety while also impacting the achievement of the United Nations Sustainable Development Goals. This study proposes a multi-scenario optimization method for urban thermal environments based on local climate zones (LCZs) in Changsha City. The research employs a genetic algorithm to optimize the LCZ quantity structure in order to improve the urban temperature environment. Subsequently, the optimized quantity structure is integrated with the future land use simulation (FLUS) model under multi-scenario constraints to achieve optimal spatial distribution of LCZs, providing scientific guidance for urban planning decision-makers. Results demonstrate that the LCZ-based optimization method can effectively regulate the urban thermal environment and maintain a suitable urban temperature range, offering both theoretical foundation and practical guidance for mitigating UHI effects.

Urbanization leads to changes in surface landscapes, such as the increase in built-up areas and the decrease in natural elements, resulting in local changes in land surface temperature, which often create unusually hot weather and affect livability, especially for mid- and low-latitude cities. Therefore, optimizing urban landscapes and adjusting the thermal environment is especially important to improve comfort and to achieve sustainable urban development. Existing studies on optimizing landscapes have considered mainly horizontal land uses/land covers but ignored their elevation. This study considered local climate zones as basic units to describe three-dimensional landscapes; we measured the relationship between local climate zones and land surface temperature, based on which the research further used a genetic algorithm and future land-use simulation models to optimize the spatial layouts of local climate zones in Changsha, China, considering multiple objectives including adjusting land surface temperature without affecting population carrying capacity, economic development, watershed protection, and forest and grass protection. According to the optimization results, the area of open low-rise buildings increased by 5.98% after optimization, and dense trees decreased by 7.64%; open low-rise buildings were suggested to be newly built in the city center and sparsely buildings should be developed in the surrounding administrative district far away from the city center. The optimization results contributed to a −5.2 °C reduction of average land surface temperature, which could significantly improve the thermal environment under the premise of ensuring the population and economic development levels and thus serves as a novel solution for improving urban landscapes to implement sustainable city development.

Building polygons plays an important role in urban management. Although leveraging deep learning techniques for building polygon extraction offers advantages, the models heavily rely on a large number of training samples to achieve good generalization performance. In scenarios with small training samples, the models struggle to effectively represent diverse building structures and handle the complexity introduced by the background. A common approach to enhance feature representation is fine-tuning a pre-trained model on a large dataset specific to the task. However, the fine-tuning process tends to overfit the model to the task area samples, leading to the loss of generalization knowledge from the large dataset. To address this challenge and enable the model to inherit the generalization knowledge from the large dataset while learning the characteristics of the task area samples, this paper proposes a knowledge distillation-based framework called Building Polygon Distillation Network (BPDNet). The teacher network of BPDNet is trained on a large building polygon dataset containing diverse building samples. The student network was trained on a small number of available samples from the target area to learn the characteristics of the task area samples. The teacher network provides guidance during the training of the student network, enabling it to learn under the supervision of generalization knowledge. Moreover, to improve the extraction of buildings against the backdrop of a complex urban context, characterized by fuzziness, irregularity, and connectivity issues, BPDNet employs the Dice Loss, which focuses attention on building boundaries. The experimental results demonstrated that BPDNet effectively addresses the problem of limited generalization by integrating the generalization knowledge from the large dataset with the characteristics of the task area samples. It accurately identifies building polygons with diverse structures and alleviates boundary fuzziness and connectivity issues.