A Sentinel-2 dataset for railway presence classification across metropolitan France

IF 1.9 Q3 MULTIDISCIPLINARY SCIENCES

Data in Brief Pub Date : 2026-08-01 Epub Date: 2026-07-23 DOI:10.1016/j.dib.2026.113104

Lukas Pasold, Luca Eisentraut, Ricardo Buettner

{"title":"A Sentinel-2 dataset for railway presence classification across metropolitan France","authors":"Lukas Pasold, Luca Eisentraut, Ricardo Buettner","doi":"10.1016/j.dib.2026.113104","DOIUrl":null,"url":null,"abstract":"<div><div>This article presents a Sentinel-2 image dataset for railway presence classification across metropolitan France. The collection contains 10,000 true-color RGB PNG chips with equal representation of railway and railway-absent scenes. Each chip has dimensions of 224 × 224 pixels and covers a ground footprint of 512 m × 512 m.</div><div>Railway chip locations were derived from retained OpenStreetMap railway=rail geometries. The railway class was constructed through geographically balanced sampling along the retained network and distance thinning between selected locations. Railway-absent chips were generated in the local surroundings of railway locations. A railway-absent chip was included only when its complete footprint did not intersect retained railway geometries and did not overlap an already accepted chip footprint.</div><div>The image collection was generated from Copernicus Sentinel-2 Level-2A observations acquired by the Sentinel-2 satellite constellation. Bands B04, B03 and B02 were used to create the red, green and blue image channels. Retrieval covered the period from 23 June 2025 to 23 June 2026 and applied a maximum cloud-cover threshold of 10 percent together with least-cloud-cover mosaicking.</div><div>The included files comprise the image collection and a CSV metadata table containing image filenames, class labels and stable candidate identifiers. The dataset provides material for railway-presence classification, comparison of image-classification architectures, spatial evaluation designs, and geospatial representation-learning workflows.</div></div>","PeriodicalId":10973,"journal":{"name":"Data in Brief","volume":"68 ","pages":"Article 113104"},"PeriodicalIF":1.9000,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Data in Brief","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2352340926006529","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2026/7/23 0:00:00","PubModel":"Epub","JCR":"Q3","JCRName":"MULTIDISCIPLINARY SCIENCES","Score":null,"Total":0}

引用次数: 0

Abstract

This article presents a Sentinel-2 image dataset for railway presence classification across metropolitan France. The collection contains 10,000 true-color RGB PNG chips with equal representation of railway and railway-absent scenes. Each chip has dimensions of 224 × 224 pixels and covers a ground footprint of 512 m × 512 m.

Railway chip locations were derived from retained OpenStreetMap railway=rail geometries. The railway class was constructed through geographically balanced sampling along the retained network and distance thinning between selected locations. Railway-absent chips were generated in the local surroundings of railway locations. A railway-absent chip was included only when its complete footprint did not intersect retained railway geometries and did not overlap an already accepted chip footprint.

The image collection was generated from Copernicus Sentinel-2 Level-2A observations acquired by the Sentinel-2 satellite constellation. Bands B04, B03 and B02 were used to create the red, green and blue image channels. Retrieval covered the period from 23 June 2025 to 23 June 2026 and applied a maximum cloud-cover threshold of 10 percent together with least-cloud-cover mosaicking.

The included files comprise the image collection and a CSV metadata table containing image filenames, class labels and stable candidate identifiers. The dataset provides material for railway-presence classification, comparison of image-classification architectures, spatial evaluation designs, and geospatial representation-learning workflows.

查看原文本刊更多论文

Sentinel-2数据集用于法国大都市铁路存在分类

本文介绍了Sentinel-2图像数据集，用于法国大都市的铁路存在分类。该系列包含10,000个真彩色RGB PNG芯片，这些芯片平等地代表了铁路和没有铁路的场景。每个芯片的尺寸为224 × 224像素，占地面积为512 m × 512 m。铁路芯片位置来自保留的OpenStreetMap铁路=铁路几何形状。铁路类是通过沿保留的网络和选定位置之间的距离细化的地理平衡采样来构建的。在铁路站点的局部环境中产生铁路缺席芯片。没有铁路的芯片只有在其完整的足迹不与保留的铁路几何形状相交并且不与已接受的芯片足迹重叠时才被包括在内。该图像采集来自Sentinel-2卫星星座获得的哥白尼Sentinel-2 Level-2A观测数据。B04， B03和B02波段用于创建红色，绿色和蓝色图像通道。反演期为2025年6月23日至2026年6月23日，最大云量阈值为10%，最小云量拼接。所包含的文件包括图像集合和包含图像文件名、类标签和稳定候选标识符的CSV元数据表。该数据集为铁路存在分类、图像分类架构比较、空间评估设计和地理空间表示学习工作流程提供了材料。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Data in Brief MULTIDISCIPLINARY SCIENCES-

CiteScore

3.10

自引率

0.00%

发文量

996

审稿时长

70 days

期刊介绍： Data in Brief provides a way for researchers to easily share and reuse each other''s datasets by publishing data articles that: -Thoroughly describe your data, facilitating reproducibility. -Make your data, which is often buried in supplementary material, easier to find. -Increase traffic towards associated research articles and data, leading to more citations. -Open up doors for new collaborations. Because you never know what data will be useful to someone else, Data in Brief welcomes submissions that describe data from all research areas.