Vectorized rooftop area data for 90 cities in China

Abstract

Reliable information on building rooftops is crucial for utilizing limited urban space effectively. In recent decades, the demand for accurate and up-to-date data on the areas of rooftops on a large-scale is increasing. However, obtaining these data is challenging due to the limited capability of conventional computer vision methods and the high cost of 3D modeling involving aerial photogrammetry. In this study, a geospatial artificial intelligence framework is presented to obtain data for rooftops using high-resolution open-access remote sensing imagery. This framework is used to generate vectorized data for rooftops in 90 cities in China. The data was validated on test samples of 180 km2 across different regions with spatial resolution, overall accuracy, and F1 score of 1 m, 97.95%, and 83.11%, respectively. In addition, the generated rooftop area conforms to the urban morphological characteristics and reflects urbanization level. These results demonstrate that the generated dataset can be used for data support and decision-making that can facilitate sustainable urban development effectively.

Geological Suitability Evaluation for Urban Underground Transportation Space Utilisation in Soft Soil Area

“…Within the range of study area, there are three canals. Figure 2(h) is the distance from existing building, which is analyzed from building rooftop area data [17]. Figure 2(i) shows the distance from existing road.…”

CBRA: The first multi-annual (2016–2021) and high-resolution (2.5 m) building rooftop area dataset in China derived with Super-resolution Segmentation from Sentinel-2 imagery

“…Typically, the ISA (Zhang et al, 2022a;Huang et al, 2022) or HSF (Marconcini et al, 2020a;Qiu et al, 2020) are derived from the imagery with a spatial resolution of decametric level (e.g., 30 m or 10 m). While the BRA (Liu et al, 2022;Zhang et al, 2022b) utilizes high-resolution aerial imagery with a resolution of metric level (e.g., 1 m). However, high-resolution aerial imagery is costly and potentially not publicly available.…”

“…To tackle it, many international endeavors analyze building rooftops by utilizing open-access Google Earth Satellite (GES) images (Liang et al, 2018). Most recently, Zhang et al (2022b) utilize GES imagery and obtained 90-cities-BRA for China under the resolution of 1 m. Due to the uneven distribution of GES image patches, as well as the inconsistent geometric offset and acquisition time (e.g., the image is obtained from various satellite sensors with different acquisition times), the existing BRA has geospatial inconsistency in practical applications, thus limiting the generalization to the questions of broad social importance, particularly in large geographic and time-scale mapping (further discussion in Sect. 2.2).…”

“…The "dual-track" society structure indicates that human activity occurs variously in both developed and developing regions of China (Guan et al, 2018a). The existing large-scale BRA dataset provided by Microsoft and Google do not include China, while the BRA produced by Zhang et al (2022b) only covers 90 cities in China. In addition, to our best knowledge, few of the existing BRA provide multi-annul results, and such temporal information is of great significance to developing countries such as China.…”

Comparing Street View Imagery and Aerial Perspectives in the Built Environment

“…In the paper, we employ a two-fold method to shed light on such comparison: a review and an experiment on extracting information about the built environment. Our literature review is systematic, while the experiments we conducted are focused on buildings since they are the predominant feature of the built environment and a topic that is frequently the subject of both perspectives (Panagiotidou et al, 2021;Ogawa et al, 2021;Wang et al, 2021;Zhang et al, 2021Zhang et al, , 2022.…”