Mapping Urban Green Spaces at the Metropolitan Level Using Very High Resolution Satellite Imagery and Deep Learning Techniques for Semantic Segmentation

Urban green spaces (UGSs) provide essential environmental services for the well-being of ecosystems and society. Due to the constant environmental, social, and economic transformations of cities, UGSs pose new challenges for management, particularly in fast-growing metropolitan areas. With technological advancement and the evolution of deep learning, it is possible to optimize the acquisition of UGS inventories through the detection of geometric patterns present in satellite imagery. This research evaluates two deep learning model techniques for semantic segmentation of UGS polygons with the use of different convolutional neural network encoders on the U-Net architecture and very high resolution (VHR) imagery to obtain updated information on UGS polygons at the metropolitan area level. The best model yielded a Dice coefficient of 0.57, IoU of 0.75, recall of 0.80, and kappa coefficient of 0.94 with an overall accuracy of 0.97, which reflects a reliable performance of the network in detecting patterns that make up the varied geometry of UGSs. A complete database of UGS polygons was quantified and categorized by types with location and delimited by municipality, allowing for the standardization of the information at the metropolitan level, which will be useful for comparative analysis with a homogenized and updated database. This is of particular interest to urban planners and UGS decision-makers.

[1]  Hui Ding,et al.  Robust liver vessel extraction using 3D U-Net with variant dice loss function , 2018, Comput. Biol. Medicine.

[2]  Collins Adjei Mensah,et al.  Destruction of urban green spaces: A problem beyond urbanization in Kumasi city (Ghana) , 2014 .

[3]  Juri Opitz,et al.  Macro F1 and Macro F1 , 2019, ArXiv.

[4]  Daniel Szatmári,et al.  Analysis of Urban Green Spaces Based on Sentinel-2A: Case Studies from Slovakia , 2017 .

[5]  Osamu Saito,et al.  Mapping Methodology of Public Urban Green Spaces Using GIS: An Example of Nagpur City, India , 2019, Sustainability.

[6]  K. Krauze,et al.  The role of informal green spaces in reducing inequalities in urban green space availability to children and seniors , 2020 .

[7]  Murat Atasoy,et al.  Monitoring the urban green spaces and landscape fragmentation using remote sensing: a case study in Osmaniye, Turkey , 2018, Environmental Monitoring and Assessment.

[8]  L. Gunderson,et al.  Adaptive governance to promote ecosystem services in urban green spaces , 2016, Urban Ecosystems.

[9]  Damir Medak,et al.  Machine learning methods for classification of the green infrastructure in city areas , 2019, ISPRS Int. J. Geo Inf..

[10]  N. Kabisch.  The Influence of Socio-economic and Socio-demographic Factors in the Association Between Urban Green Space and Health , 2019, Biodiversity and Health in the Face of Climate Change.

[11]  Luc Van Gool,et al.  Deep Retinal Image Understanding , 2016, MICCAI.

[12]  Baofeng Su,et al.  Significant Remote Sensing Vegetation Indices: A Review of Developments and Applications , 2017, J. Sensors.

[13]  E. Gómez‐Baggethun,et al.  Classifying and valuing ecosystem services for urban planning , 2013 .

[14]  A. Huete,et al.  A review of vegetation indices , 1995 .

[15]  K. Rahman,et al.  Analyzing the Level of Accessibility of Public Urban Green Spaces to Different Socially Vulnerable Groups of People , 2018, Sustainability.

[16]  Zhiyu Xu,et al.  A Novel Intelligent Classification Method for Urban Green Space Based on High-Resolution Remote Sensing Images , 2020, Remote. Sens..

[17]  D. Hochuli,et al.  Defining greenspace: Multiple uses across multiple disciplines , 2017 .

[18]  J. Klopp,et al.  The urban sustainable development goal: Indicators, complexity and the politics of measuring cities , 2017 .

[19]  Jing Zhang,et al.  Assessing the Distribution of Urban Green Spaces and its Anisotropic Cooling Distance on Urban Heat Island Pattern in Baotou, China , 2016, ISPRS Int. J. Geo Inf..

[20]  Cristina Branquinho,et al.  Modeling the provision of air-quality regulation ecosystem service provided by urban green spaces using lichens as ecological indicators. , 2019, The Science of the total environment.

[21]  Gerd Weitkamp,et al.  Quality over Quantity: Contribution of Urban Green Space to Neighborhood Satisfaction , 2017, International journal of environmental research and public health.

[22]  The state and use of municipal tree inventories in Swedish municipalities – results from a national survey , 2018, Urban Ecosystems.

[23]  Ben Somers,et al.  Mapping Functional Urban Green Types Using High Resolution Remote Sensing Data , 2020, Sustainability.

[24]  Natalia Gimelshein,et al.  PyTorch: An Imperative Style, High-Performance Deep Learning Library , 2019, NeurIPS.

[25]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[26]  Mojca Nastran,et al.  Urban heat islands in relation to green land use in European cities , 2019, Urban Forestry & Urban Greening.

[27]  Shuguang Liu,et al.  Spatiotemporal trends of terrestrial vegetation activity along the urban development intensity gradient in China's 32 major cities. , 2014, The Science of the total environment.

[28]  Bruno Martins,et al.  Index for evaluation of public parks and gardens proximity based on the mobility network: A case study of Braga, Braganza and Viana do Castelo (Portugal) and Lugo and Pontevedra (Spain) , 2018, Urban Forestry & Urban Greening.

[29]  Jacob Cohen.  A Coefficient of Agreement for Nominal Scales , 1960 .

[30]  Michael W. Strohbach,et al.  The ‘GartenApp’: Assessing and Communicating the Ecological Potential of Private Gardens , 2019 .

[31]  Jeremy Howard,et al.  fastai: A Layered API for Deep Learning , 2020, Inf..

[32]  T. Gillespie,et al.  Monitoring changes of NDVI in protected areas of southern California , 2018 .

[33]  Fengdi Ma.  Spatial equity analysis of urban green space based on spatial design network analysis (sDNA): A case study of central Jinan, China , 2020 .

[34]  Pawan Gupta,et al.  Spatial and Temporal Distribution of PM2.5 Pollution over Northeastern Mexico: Application of MERRA-2 Reanalysis Datasets , 2020, Remote. Sens..

[35]  Hiram Calvo,et al.  Deep Green Diagnostics: Urban Green Space Analysis Using Deep Learning and Drone Images , 2019, Sensors.

[36]  H. Skokanová,et al.  Mapping Green Infrastructure Elements Based on Available Data, A Case Study of the Czech Republic , 2020 .

[37]  Ying Zhang,et al.  Application of the Dice Coefficient to Accuracy Assessment of Object-Based Image Classification , 2017 .

[38]  Yongyang Xu,et al.  Road Extraction from High-Resolution Remote Sensing Imagery Using Deep Learning , 2018, Remote. Sens..

[39]  Rebecca B. Dikow,et al.  Generating segmentation masks of herbarium specimens and a data set for training segmentation models using deep learning , 2020, Applications in plant sciences.

[40]  A. Sharifi,et al.  The Effect of Rapid Population Growth on Urban Expansion and Destruction of Green Space in Tehran from 1972 to 2017 , 2019, Journal of the Indian Society of Remote Sensing.

[41]  Denise R. Aberle,et al.  An Interpretable Deep Hierarchical Semantic Convolutional Neural Network for Lung Nodule Malignancy Classification , 2018, Expert Syst. Appl..

[42]  J. Wolch,et al.  Urban green space, public health, and environmental justice: The challenge of making cities ‘just green enough’ , 2014 .

[43]  L. R. Dice.  Measures of the Amount of Ecologic Association Between Species , 1945 .

[44]  Arijit Roy,et al.  GIS based analysis for assessing the accessibility at hierarchical levels of urban green spaces , 2016 .

[45]  Biplab Banerjee,et al.  Machine learning on high performance computing for urban greenspace change detection: satellite image data fusion approach , 2020 .

[46]  Hao Wu,et al.  Evaluation and Planning of Urban Green Space Distribution Based on Mobile Phone Data and Two-Step Floating Catchment Area Method , 2018 .

[47]  Marciano Saraiva,et al.  Automatic Mapping of Center Pivot Irrigation Systems from Satellite Images Using Deep Learning , 2020, Remote. Sens..

[48]  Monika Kuffer,et al.  Cloud Computation Using High-Resolution Images for Improving the SDG Indicator on Open Spaces , 2020, Remote. Sens..

[49]  J. Banegas,et al.  Impact of road traffic noise on cause-specific mortality in Madrid (Spain). , 2017, The Science of the total environment.

[50]  D. Lozano-García,et al.  Precision of remote sensors to estimate aerial biomass parameters: portable LIDAR and optical sensors , 2018 .

[51]  Derya Gülçin,et al.  Mapping Urban Green Spaces Based on an Object-Oriented Approach , 2018, Bilge International Journal of Science and Technology Research.

[52]  Aapo Huovila,et al.  Benchmarking Nature-Based Solution and Smart City Assessment Schemes Against the Sustainable Development Goal Indicator Framework , 2018, Front. Environ. Sci..

[53]  J. Ndambuki,et al.  Accuracy Assessment of Land Use/Land Cover Classification Using Remote Sensing and GIS , 2017 .

[54]  Xiao Xiang Zhu,et al.  Semantic segmentation of slums in satellite images using transfer learning on fully convolutional neural networks , 2019, ISPRS Journal of Photogrammetry and Remote Sensing.

[55]  C. Berlanga-Robles,et al.  Spatio-temporal mangrove canopy variation (2001–2016) assessed using the MODIS enhanced vegetation index (EVI) , 2019, Journal of Coastal Conservation.

[56]  Qingsong He,et al.  Dismantling the fence for social justice? Evidence based on the inequity of urban green space accessibility in the central urban area of Beijing , 2018, Environment and Planning B: Urban Analytics and City Science.

[57]  Yan Liu,et al.  Deep residual learning for image steganalysis , 2018, Multimedia Tools and Applications.

[58]  Zylshal,et al.  A support vector machine object based image analysis approach on urban green space extraction using Pleiades-1A imagery , 2016, Modeling Earth Systems and Environment.

[59]  Xiaodong Zhu,et al.  Quantifying landscape-metrics impacts on urban green-spaces and water-bodies cooling effect: The study of Nanjing, China , 2020 .

[60]  Jagannath Aryal,et al.  Mapping Urban Tree Cover Changes Using Object-Based Convolution Neural Network (OB-CNN) , 2020, Remote. Sens..

[61]  Hod Lipson,et al.  Quantitative Phenotyping of Northern Leaf Blight in UAV Images Using Deep Learning , 2019, Remote. Sens..

[62]  Elizabeth A. Corley,et al.  Green spaces and heterogeneous social groups in the U.S. , 2020 .