{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T14:41:15Z","timestamp":1781620875576,"version":"3.54.5"},"reference-count":31,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,8,13]],"date-time":"2025-08-13T00:00:00Z","timestamp":1755043200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Big Data"],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>OpenStreetMap (OSM) road surface data is critical for navigation, infrastructure monitoring, and urban planning but is often incomplete or inconsistent. This study addresses the need for automated validation and classification of road surfaces by leveraging high-resolution aerial imagery and deep learning techniques.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>We propose a MaskCNN-based deep learning model enhanced with attention mechanisms and a hierarchical loss function to classify road surfaces into four types: asphalt, concrete, gravel, and dirt. The model uses NAIP (National Agriculture Imagery Program) aerial imagery aligned with OSM labels. Preprocessing includes georeferencing, data augmentation, label cleaning, and class balancing. The architecture comprises a ResNet-50 encoder with squeeze-and-excitation blocks and a U-Net-style decoder with spatial attention. Evaluation metrics include accuracy, mIoU, precision, recall, and F1-score.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>The proposed model achieved an overall accuracy of 92.3% and a mean Intersection over Union (mIoU) of 83.7%, outperforming baseline models such as SVM (81.2% accuracy), Random Forest (83.7%), and standard U-Net (89.6%). Class-wise performance showed high precision and recall even for challenging surface types like gravel and dirt. Comparative evaluations against state-of-the-art models (COANet, SA-UNet, MMFFNet) also confirmed superior performance.<\/jats:p><\/jats:sec><jats:sec><jats:title>Discussion<\/jats:title><jats:p>The results demonstrate that combining NAIP imagery with attention-guided CNN architectures and hierarchical loss functions significantly improves road surface classification. The model is robust across varied terrains and visual conditions and shows potential for real-world applications such as OSM data enhancement, infrastructure analysis, and autonomous navigation. Limitations include label noise in OSM and class imbalance, which can be addressed through future work involving semi-supervised learning and multimodal data integration.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fdata.2025.1657320","type":"journal-article","created":{"date-parts":[[2025,8,13]],"date-time":"2025-08-13T14:56:08Z","timestamp":1755096968000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["Automated road surface classification in OpenStreetMap using MaskCNN and aerial imagery"],"prefix":"10.3389","volume":"8","author":[{"given":"R.","family":"Parvathi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"V.","family":"Pattabiraman","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nancy","family":"Saxena","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aakarsh","family":"Mishra","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Utkarsh","family":"Mishra","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ansh","family":"Pandey","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,8,13]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"1444","DOI":"10.3390\/rs12091444","article-title":"Deep learning approaches applied to remote sensing datasets for road extraction: a state-of-the-art review","volume":"12","author":"Abdollahi","year":"2020","journal-title":"Remote Sens."},{"key":"B2","doi-asserted-by":"publisher","first-page":"e0180099","DOI":"10.1371\/journal.pone.0180698","article-title":"The world's open-source mapping platform: OpenStreetMap","volume":"12","author":"Barrington-Leigh","year":"2017","journal-title":"PLoS ONE"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-35289-8_26","article-title":"Practical recommendations for gradient-based training","author":"Bengio","year":"2012","journal-title":"arXiv"},{"key":"B4","first-page":"1784","article-title":"Attention-based deep learning for road surface classification using remote sensing data","volume":"24","author":"Chen","year":"2023","journal-title":"IEEE Transact. Intell. Transport. Syst."},{"key":"B5","volume-title":"Deep Learning with Python","author":"Chollet","year":"2017"},{"key":"B6","doi-asserted-by":"publisher","first-page":"72","DOI":"10.5194\/isprs-annals-V-3-2020-669-2020","article-title":"Mapping road surfaces using OpenStreetMap and remote sensing data","volume":"9","author":"Fonte","year":"2020","journal-title":"ISPRS Int. J. Geoinf."},{"key":"B7","doi-asserted-by":"publisher","first-page":"682","DOI":"10.1068\/b35097","article-title":"How good is volunteered geographical information? A comparative study of OpenStreetMap and Ordnance Survey datasets","volume":"37","author":"Haklay","year":"2010","journal-title":"Environ. Plann. B Plann. Des."},{"key":"B8","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90","article-title":"Deep residual learning for image recognition","author":"He","year":"2016","journal-title":"Comp. Vis. Pattern Recogn"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00745","article-title":"Squeeze-and-excitation networks","author":"Hu","year":"2018","journal-title":"Comp. Vis. Pattern Recogn"},{"key":"B10","doi-asserted-by":"publisher","first-page":"a004868","DOI":"10.1093\/oxfordjournals.pan.a004868","article-title":"Logistic regression in rare events data","volume":"9","author":"King","year":"2001","journal-title":"Polit. Anal."},{"key":"B11","first-page":"1246","article-title":"\u201cDeep hierarchical semantic segmentation,\u201d","author":"Li","year":"2022","journal-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.324","article-title":"Focal loss for dense object detection","author":"Lin","year":"2017","journal-title":"ICCV"},{"key":"B13","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1608.03983","article-title":"\u201cSGDR: stochastic gradient descent with warm restarts,\u201d","author":"Loshchilov","year":"2017","journal-title":"ICLR"},{"key":"B14","doi-asserted-by":"publisher","first-page":"645","DOI":"10.1109\/TGRS.2016.2612821","article-title":"Convolutional neural networks for large-scale remote sensing image classification","volume":"55","author":"Maggiori","year":"2017","journal-title":"IEEE Transact. Geosci. Remote Sens."},{"key":"B15","doi-asserted-by":"publisher","first-page":"1205","DOI":"10.1109\/TGRS.2016.2612821","article-title":"Improving OpenStreetMap road classification using remote sensing data","volume":"11","author":"Maggiori","year":"2019","journal-title":"Remote Sens."},{"key":"B16","doi-asserted-by":"publisher","first-page":"8540","DOI":"10.1109\/TIP.2021.3117076","article-title":"COANet: connectivity attention network for road extraction from satellite imagery","volume":"30","author":"Mei","year":"2021","journal-title":"IEEE Transact. Image Process."},{"key":"B17","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1710.03740","article-title":"\u201cMixed precision training,\u201d","author":"Micikevicius","year":"2018","journal-title":"ICLR."},{"key":"B18","doi-asserted-by":"publisher","first-page":"210","DOI":"10.1007\/978-3-642-15567-3_16","article-title":"Learning to detect roads in high-resolution aerial images","volume":"6316","author":"Mnih","year":"2010","journal-title":"Eur. Conf. Comp. Vis."},{"key":"B19","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1804.03999","article-title":"Attention U-Net: learning where to look for the pancreas","author":"Oktay","year":"2018","journal-title":"arXiv"},{"key":"B20","doi-asserted-by":"publisher","first-page":"55","DOI":"10.1007\/3-540-49430-8_3","article-title":"Early stopping - but when?","volume":"1524","author":"Prechelt","year":"1998","journal-title":"Neur. Netw."},{"key":"B21","doi-asserted-by":"publisher","first-page":"234","DOI":"10.1007\/978-3-319-24574-4_28","article-title":"U-Net: convolutional networks for biomedical image segmentation","volume":"9351","author":"Ronneberger","year":"2015","journal-title":"Med. Image Comp. Comp. Assist. Interv."},{"key":"B22","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298965","article-title":"Fully convolutional networks for semantic segmentation","author":"Shelhamer","year":"2015","journal-title":"IEEE Conf. Comp. Vis. Patt. Recognit."},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1606.02585","article-title":"Fully convolutional networks for dense semantic labelling of high-resolution aerial imagery","author":"Sherrah","year":"2016","journal-title":"arXiv"},{"key":"B24","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1186\/s40537-019-0197-0","article-title":"A survey on image data augmentation for deep learning","volume":"6","author":"Shorten","year":"2019","journal-title":"J. Big Data."},{"key":"B25","doi-asserted-by":"crossref","first-page":"605","DOI":"10.1007\/978-981-99-0969-8_63","article-title":"\u201cConsistency assessment of OpenStreetMap road dataset of Haryana and Punjab using K-means and elbow method,\u201d","volume-title":"Signals, Machines and Automation","author":"Singh","year":"2023"},{"key":"B26","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TGRS.2024.3379988","article-title":"A multiscale and multidirection feature fusion network for road detection from satellite imagery","volume":"62","author":"Wang","year":"2024","journal-title":"IEEE Transact. Geosci. Remote Sens."},{"key":"B27","article-title":"SA-UNet: spatial attention U-Net for road segmentation in remote sensing imagery","author":"Wu","year":"2024","journal-title":"IEEE Transact. Geosci. Remote Sens"},{"key":"B28","first-page":"1775","article-title":"Hierarchical deep learning for road surface segmentation","volume":"29","author":"Xu","year":"2018","journal-title":"IEEE Transact. Neur. Netw. Learn. Syst."},{"key":"B29","doi-asserted-by":"publisher","first-page":"1461","DOI":"10.3390\/rs10091461","article-title":"Road extraction from high-resolution remote sensing imagery using deep learning","volume":"10","author":"Zhang","year":"2018","journal-title":"Remote Sens."},{"key":"B30","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TGRS.2021.3093474","article-title":"Deep learning-based road surface classification using aerial imagery and crowdsourced data","volume":"60","author":"Zhang","year":"2022","journal-title":"IEEE Transact. Geosci. Remote Sens."},{"key":"B31","first-page":"2916","article-title":"Road classification from high-resolution aerial imagery using deep learning","volume":"55","author":"Zhang","year":"2017","journal-title":"IEEE Transact. Geosci. Remote Sens."}],"container-title":["Frontiers in Big Data"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fdata.2025.1657320\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,13]],"date-time":"2025-08-13T14:56:10Z","timestamp":1755096970000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fdata.2025.1657320\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,13]]},"references-count":31,"alternative-id":["10.3389\/fdata.2025.1657320"],"URL":"https:\/\/doi.org\/10.3389\/fdata.2025.1657320","relation":{},"ISSN":["2624-909X"],"issn-type":[{"value":"2624-909X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,8,13]]},"article-number":"1657320"}}