| 296 | 0 | 816 |
| 下载次数 | 被引频次 | 阅读次数 |
结合多源地理空间数据进行城市细分地类的识别是城市用地监测的重要手段，对提升城市管理的智慧化水平具有重要意义。该文以武汉市中心城区为例，基于Vision-Transformer模型和多源数据融合策略对城市地类进行识别，在此基础上，引入注意力归因得分方法计算兴趣点(POI)对模型识别结果的贡献度，并采用核密度分析探究城市细分地类与人类活动间的关系。结果表明：POI的空间分异与类别组合差异所提供的不同维度特征能将城市地类识别精度提高至93.0%;不同地类中包含的POI组合与地类的功能关联较强，高集聚度的餐饮、购物类POI代表商业/服务业用地性质，孤立、稀疏的居民楼代表居住用地性质，医疗、科教类建筑及周边较低集聚度的餐饮、购物类POI代表公共服务用地性质。
Abstract：Identification of urban subdivided land use types based on multi-source geographic spatial data serves as a vital tool for urban land use monitoring, which is of great significance for enhancing smart urban governance.This study takes the central urban area of Wuhan as an example, and employs the Vision Transformer model with a multi-source data fusion strategy to identify urban land use types.Furthermore, an attention attribution score method is introduced to quantitatively assess the contribution of point of interest(POI) data to the model recognition results.Kernel density analysis is also applied to investigate the relationships between subdivided land use types and human activities.The experimental results demonstrate that the spatial distribution differences and category combination differences of POIs provide multi-dimensional features, which improve urban land use identification accuracy to 93.00%.Moreover, the combinations of POIs contained in different land use types are strongly correlated with their inherent functions.Specifically, high concentration of catering and shopping POIs is indicative of commercial/service land use, whereas isolated and sparse residential buildings, medical, and science-education buildings, along with the lower-density catering and shopping POIs in their vicinity, are characteristic of residential and public service land uses, respectively.
[1] RAMAN R,ROY U K.Taxonomy of urban mixed land use planning[J].Land Use Policy,2019,88:104102.DOI:10.1016/j.landusepol.2019.104102.
[2] ZHENG H Y,ZHUO Y F,XU Z G,et al.Measuring and characterizing land use mix patterns of China′s megacities:a case study of Shanghai[J].Growth and Change,2021,52(4):2509-2539.
[3] LV P Y,WU W J,ZHONG Y F,et al.SCViT:a spatial-channel feature preserving vision transformer for remote sensing image scene classification[J].IEEE Transactions on Geoscience and Remote Sensing,2022,60:1-12.
[4] PAN S J,YANG Q.A survey on transfer learning[J].IEEE Transactions on Knowledge and Data Engineering,2010,22(10):1345-1359.
[5] ELWOOD S,GOODCHILD M F,SUI D Z.Researching volunteered geographic information:spatial data,geographic research,and new social practice[J].Annals of the Association of American Geographers,2012,102(3):571-590.
[6] GRANT M,KRZYSZTOF J,SONG G,et al.POI pulse:a multigranular,semantic signatures-based information observatory for the interactive visualization of big geosocial data[J].Cartographica,2015,50(2):71-85.
[7] 池娇，焦利民，董婷，等.基于POI数据的城市功能区定量识别及其可视化[J].测绘地理信息，2016,41(2):68-73.
[8] PSYLLIDIS A,GAO S,HU Y J,et al.Points of interest (POI):a commentary on the state of the art,challenges,and prospects for the future[J].Computational Urban Science,2022,2(1):20.DOI:10.1007/s43762-022-00047-w.
[9] LIU W,WU W J,THAKURIAH P,et al.The geography of human activity and land use:a big data approach[J].Cities,2020,97:102523.DOI:10.1016/j.cities.2019.102523.
[10] LIU X P,HE J L,YAO Y,et al.Classifying urban land use by integrating remote sensing and social media data[J].International Journal of Geographical Information Science,2017,31(8):1675-1696.
[11] WU H,LIN A Q,XING X D,et al.Identifying core driving factors of urban land use change from global land cover products and POI data using the random forest method[J].International Journal of Applied Earth Observation and Geoinformation,2021,103:102475.DOI:10.1016/j.jag.2021.102475.
[12] GUO Y W,TANG J B,LIU H M,et al.Identifying up-to-date urban land-use patterns with visual and semantic features based on multisource geospatial data[J].Sustainable Cities and Society,2024,101:105184.DOI:10.1016/j.scs.2024.105184.
[13] ZHOU W,PERSELLO C,LI M M,et al.Building use and mixed-use classification with a transformer-based network fusing satellite images and geospatial textual information[J].Remote Sensing of Environment,2023,297:113767.DOI:10.1016/j.rse.2023.113767.
[14] YAO Y,YAN X Q,LUO P,et al.Classifying land-use patterns by integrating time-series electricity data and high-spatial resolution remote sensing imagery[J].International Journal of Applied Earth Observation and Geoinformation,2022,106:102664.DOI:10.1016/j.jag.2021.102664.
[15] YAN X Q,JIANG Z W,LUO P,et al.A multimodal data fusion model for accurate and interpretable urban land use mapping with uncertainty analysis[J].International Journal of Applied Earth Observation and Geoinformation,2024,129:103805.DOI:10.1016/j.jag.2024.103805.
[16] WANG L,LI Z M,ZHANG Z.City profile:Wuhan 2004-2020[J].Cities,2022,123:103585.DOI:10.1016/j.cities.2022.103585.
[17] HAN K,WANG Y H,CHEN H T,et al.A survey on vision transformer[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45(1):87-110.
[18] RAGHU M,UNTERTHINER T,KORNBLITH S.Do vision transformers see like convolutional neural networks[J].Advances in Neural Information Processing Systems,2021,34:12116-12128.
[19] DENG J,DONG W,SOCHER R,et al.ImageNet:a large-scale hierarchical image database[C]//2009 IEEE Conference on Computer Vision and Pattern Recognition.Miami:IEEE,2009:248-255.
[20] VASWANI A,SHAZEER N,PARMAR N,et al.Attention is all you need[J].Advances in Neural Information Processing Systems,2017,30:5999-6009.
[21] ABNAR S,ZUIDEMA W.Quantifying attention flow in transformers[EB/OL].(2020-05-31)[2025-07-18].https://arxiv.org/abs/2005.00928.
[22] XU L,YAN X,DING W Y,et al.Attribution rollout:a new way to interpret visual transformer[J].Journal of Ambient Intelligence and Humanized Computing,2023,14(1):163-173.
[23] JAIN S,WALLACE B C.Attention is not explanation[C]//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies,Volume 1 (Long and Short Papers).Minneapolis:Association for Computational Linguistics,2019:3543-3556.
基本信息:
DOI：
中图分类号:P208;TU984.113
引用信息:
[1]胡伟,刘嘉宇,周琛.基于多源数据融合的城市细分地类识别与可解释性分析[J].地理与地理信息科学,2025,41(06):26-32.
基金信息:
国家自然科学基金项目（42271414）; 江苏省自然科学基金项目（BK20231410）