Ma Yibin, Chen Pengfei, Gong Mengjie, Cai Yixian, Jian Izzy Yi
School of Geospatial Engineering and Science, Sun Yat-sen University, and Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai, 519082, China.
Key Laboratory of Comprehensive Observation of Polar Environment (Sun Yat-sen University), Ministry of Education, Zhuhai, 519082, China.
Sci Data. 2025 Aug 5;12(1):1356. doi: 10.1038/s41597-025-05706-1.
Multi-temporal mapping of the Green View Index (GVI) is crucial for understanding how urban residents perceive seasonal changes in streetscape greenness. Compared to street view imagery (SVI), remote sensing data offers higher temporal frequency and broader spatial coverage, enabling large-scale dynamic monitoring. However, most existing GVI estimation methods rely heavily on SVI, limiting their ability to support cross-city and seasonal analysis. To address this gap, we present the Seasonal Green View Index 2023 (SGVI-2023), a GVI mapping dataset derived from multisource remote sensing data and deep learning. Covering 19 major Chinese cities, SGVI-2023 was developed using approximately 1 million paired samples of satellite and SVI data collected from 2019 to 2023. All data underwent strict preprocessing and partitioning. Evaluation results show strong accuracy, with Pearson correlations of 0.867 at the point scale and 0.918 at the street scale. As the first cross-city, seasonally resolved GVI dataset based on remote sensing, SGVI-2023 provides valuable support for human-centered urban greenness monitoring and data-driven urban planning.
绿色视野指数（GVI）的多时相映射对于理解城市居民如何感知街道景观绿化的季节变化至关重要。与街景图像（SVI）相比，遥感数据具有更高的时间频率和更广泛的空间覆盖范围，能够进行大规模动态监测。然而，大多数现有的GVI估计方法严重依赖SVI，限制了它们支持跨城市和季节分析的能力。为了填补这一空白，我们提出了2023年季节性绿色视野指数（SGVI-2023），这是一个从多源遥感数据和深度学习中得出的GVI映射数据集。SGVI-2023覆盖了中国19个主要城市，它是利用2019年至2023年收集的约100万对卫星和SVI数据样本开发的。所有数据都经过了严格的预处理和划分。评估结果显示出很高的准确性，在点尺度上的皮尔逊相关系数为0.867，在街道尺度上为0.918。作为首个基于遥感的跨城市、按季节解析的GVI数据集，SGVI-2023为以人类为中心的城市绿化监测和数据驱动的城市规划提供了有价值的支持。
