[转载]深度学习中的遥感影像数据集
收集网络上公开的遥感数据集，欢迎补充
Ref：
https://blog.csdn.net/weixin_36670529/article/details/84580431?depth_1-utm_source=distribute.pc_relevant.none-task&utm_source=distribute.pc_relevant.none-task
https://blog.csdn.net/dangdangliu/article/details/104323166
http://www.graphnetcloud.cn/page89
https://mp.weixin.qq.com/s/weL3hqqBD0P-donV2an2yw
https://mp.weixin.qq.com/s/xU9zYTMdhUgmcuMUSuhzpw
https://zhangbin0917.github.io/2018/06/12/%E9%81%A5%E6%84%9F%E6%95%B0%E6%8D%AE%E9%9B%86/
http://homepages.inf.ed.ac.uk/rbf/CVonline/Imagedbase.htm#remote
注：
- 以下数据集均为网上开源数据集，多数数据集可访问、且已下载或保存下载链接
- 若部分链接不可访问，可以评论或者留言
- *）星号 表示数据质量未知
删除线表示连接暂时无法访问；- 本片博客直接引用深度学习的遥感影像数据集，感谢分享，若有侵权，联系自删！原作者说会持续更新，盆友们记得关注~
目录
13）Planet: Understanding the Amazon from Space
10）Planet: Understanding the Amazon from Space
15）Airbus Ship Detection Challenge
*）Dstl Satellite Imagery Feature Detection
*)2018 Open AI Tanzania Building Footprint Segmentation Challenge
*）Functional Map of the World Challenge
*)Airbus Ship Detection Challenge
*）MASATI dataset - MAritime SATellite Imagery dataset
*）MAFAT Challenge - Fine-Grained Classification of Objects from Aerial Imagery
3）ISPRS 2D Semantic Labeling Contest
4）DeepGlobe Land Cover Classification Challenge
*)Slovenia 2017 Land Cover Classification Dataset
1）Road and Building Detection Datasets
3）38-Cloud: A Cloud Segmentation Dataset
2.Satellite dataset I (global cities)
3. Satellite dataset Ⅱ (East Asia)
*）Massachusetts Road and Building Detection Datasets
Massachusetts Buildings Dataset
*)Aerial Image Segmentation Dataset
2）SZTAKI AirChange Benchmark Set
6）WHU ： Building change detection dataset
7）Season-varying Change Detection Dataset
8）Onera Satellite Change Detection
*）AIST Building Change Detection
*）Urban Atlas Change 2012-2018
*）High Resolution Change Detection
影像分类
1）UC Merced Land-Use Data Set
Size：256*256；
Num：21classes*100=2100pic
http://weegee.vision.ucmerced.edu/datasets/landuse.html
2）WHU-RS19
Size：600*600；
Num：19classes*50±≈1000pic
http://captain.whu.edu.cn/repository.html
http://captain.whu.edu.cn/datasets/WHU-RS19.zip
3）AID
Size：600*600；
Num：30classes*300±≈10000pic
http://captain.whu.edu.cn/repository.html
https://captain-whu.github.io/AID/
https://pan.baidu.com/s/1mifOBv6#list/path=%2F
4）SIRI-WHU Data Set
Size：200*200；
Num：12classes*200=2400pic
http://www.lmars.whu.edu.cn/prof_web/zhongyanfei/e-code.html
5）RSSCN7 Data Set
Size：400*400；
Num：7classes*400=2400pic
https://sites.google.com/site/qinzoucn/documents
https://pan.baidu.com/s/1slSn6Vz
6）NWPU-RESISC45 dataset
Size：256*256
Num：45classes*700=31500pic
http://www.escience.cn/people/JunweiHan/NWPU-RESISC45.html
https://pan.baidu.com/s/1mifR6tU#list/path=%2F
7）OPTIMAL-31
Size：256*256
Num：10class*2700±≈1860pic
http://crabwq.github.io/
https://drive.google.com/open?id=1Fk9a0DW8UyyQsR8dP2Qdakmr69NVBhq9
8）BigEarthNet
Num：590326 image patches，each patch include
Size：120*120，10m bands
60*60，20m bands
20*20，60m bands
http://bigearth.net/
9）RSI-CB
RSI-CB128
Size：128×128
Num：45classes*800±≈36000pic
https://github.com/lehaifeng/RSI-CB
https://pan.baidu.com/s/1pLnZQ23
RSI-CB256
Size：256×256
Num：35classes*690±≈24000pic
https://github.com/lehaifeng/RSI-CB
https://pan.baidu.com/s/1bpIQ0IN
10）RSSCN7 Dataset
Size：400*400
Num：7classes，2800pic；
https://pan.baidu.com/s/1slSn6Vz
https://sites.google.com/site/qinzoucn/documents/readme_RSSCN7.pdf?attredirects=0&d=1
https://sites.google.com/site/qinzoucn/documents
11）EuroSAT
Size：64*64
Num：31class*60=27000pic
https://github.com/phelber/eurosat#
https://www.kesci.com/home/dataset/5e6b3125dd480d002c21c46c
12）PatternNet
Size：256*256
Num：38classes*800=30400pic
https://sites.google.com/view/zhouwx/dataset#h.p_Tgef10WTuEFr
https://drive.google.com/file/d/127lxXYqzO6Bd0yZhvEbgIfz95HaEnr9K/view?usp=sharing
13）Planet: Understanding the Amazon from Space
Num：17class
https://www.kaggle.com/c/planet-understanding-the-amazon-from-space/data
https://www.kesci.com/home/dataset/5dfc38662823a10036ada832
14）SAT-4 and SAT-6
Size：28*28
Num：4classes，500000pic；6classes，405000pic
http://csc.lsu.edu/~saikat/deepsat/
https://drive.google.com/uc?id=0B0Fef71_vt3PUkZ4YVZ5WWNvZWs&export=download
15）RS_C11
Size：512*512
Num： 11 classes，1232pic
https://www.researchgate.net/publication/271647282_RS_C11_Database/comments
https://pan.baidu.com/s/1mhagndY
https://hyper.ai/datasets/5443
目标检测
1）DOTA， DOAI
v1.0 DOTA
Num：15classes，2806pic，188282instances
https://captain-whu.github.io/DOTA/index.html
http://captain.whu.edu.cn/DOTAweb/dataset.html
v1.5 DOAI
Num：16classes，2806pic，400000±instances
http://captain.whu.edu.cn/DOTAweb/
https://captain-whu.github.io/DOAI2019/dataset.html
2）TGRS-HRRSD-Dataset
Num：13classes，21761pic，55740instances
https://github.com/CrazyStoneonRoad/TGRS-HRRSD-Dataset
3）NWPU VHR-10 dataset
Num：10classes，800pic
http://www.escience.cn/people/JunweiHan/NWPUVHR10dataset.html
https://pan.baidu.com/s/1hqwzXeG#list/path=%2F
4）DIOR dataset
Num：23463pic，190288instances
http://www.escience.cn/people/JunweiHan/DIOR.html
https://pan.baidu.com/s/1iLKT0JQoKXEJTGNxt5lSMg#list/path=%2F
5）xView
Num：60classes，1000000±instances
http://xviewdataset.org/
https://challenge.xviewdataset.org/data-download
6）RSOD-Dataset
Num：
aircraft：4993 aircrafts in 446 images.
playground： 191 playgrounds in 189 images.
overpass： 180 overpass in 176 overpass.
oiltank：1586 oiltanks in 165 images.
https://github.com/RSIA-LIESMARS-WHU/RSOD-Dataset-
7）UCAS-AOD
Num：
car：7114 cars in 510 images.
plane： 7482 plane in 1000 images.
https://ucassdl.cn/
https://onedrive.hyper.ai/home/UCAS-AOD/data
8）OIRDS
Num：1classes(car)，900pic，1800instances
https://sourceforge.net/projects/oirds/
9）VEDAI
Num：3classes(cars)，982pic，2349instances
https://downloads.greyc.fr/vedai/
10）Planet: Understanding the Amazon from Space
Size：128*128
Num：
damage build,，(10000+90000+2000+2000)pic
no damage build
https://www.kaggle.com/kmader/satellite-images-of-hurricane-damage
https://www.kesci.com/home/dataset/5da6e8a6c83fb40042073e4f/document
11）LEVIR
Size：800 * 600
Num：3class，22000pic，11000instances
http://levir.buaa.edu.cn/Code.htm
https://pan.baidu.com/s/1geTwAVD
12）COWC
Num：
car， 32716instances
negative， 58247instances
https://gdo152.llnl.gov/cowc/
https://gdo152.llnl.gov/cowc/download/
13）HRSC2016
Num：1class(ship)，(436+181+444)pic，(1207+541+1228)instances
https://sites.google.com/site/hrsc2016/
http://www.escience.cn/people/liuzikun/DataSet.html
14）AIR-SARShip-1.0
Num：1class(ship)，31pic，461instances
http://radars.ie.ac.cn/web/data/getData?dataType=SARDataset
15）Airbus Ship Detection Challenge
Num：1class(ship)，192556pic，
https://blog.csdn.net/weixin_42880443/article/details/82015347
链接: https://pan.baidu.com/s/1RbGPNr05orfEbUA0u9CmJg 密码: p9b8
16）SAR-Ship-Dataset
Size：256*256
Num：1class(ship)，43,819pic，
https://github.com/CAESAR-Radi/SAR-Ship-Dataset
17）CrowdAI Mapping Challenge
Size：300*300
class：build，（280741+60317）pic
https://www.crowdai.org/challenges/mapping-challenge
https://www.jianshu.com/p/90efc39975da
https://pan.baidu.com/s/1wm4WtOLGiknaqAJSrlzufA 密码:xho2
*）ITCVD
Num：1class(car)，(135+38)pic，(23543+5545)instances
https://eostore.itc.utwente.nl:5001/fsdownload/zZYfgbB2X/ITCVD
链接：https://pan.baidu.com/s/1S77GCYBpAGjD-BSQjpKJ8g 提取码：kzlh
*）Dstl Satellite Imagery Feature Detection
Num：10class
https://www.kaggle.com/c/dstl-satellite-imagery-feature-detection/data
*)2018 Open AI Tanzania Building Footprint Segmentation Challenge
https://docs.google.com/spreadsheets/d/1tP133OqpwvkzHnkmS_3nezpTJMq06tpPM6P2qU6kaZ4/edit?usp=sharing
https://competitions.codalab.org/competitions/20100
*）Functional Map of the World Challenge
Num：63class，31pic，1000000instances
https://www.iarpa.gov/challenges/fmow.html
*)Airbus Ship Detection Challenge
https://www.kaggle.com/c/airbus-ship-detection
*)OpenSARShip
http://opensar.sjtu.edu.cn/
*）ccmData2008
（忘记了是从哪里下载的，该名称为压缩包名称）
Num：425pic
https://download.csdn.net/download/nominior/12341007
*）MASATI dataset - MAritime SATellite Imagery dataset
Size：512*512
Num：6classes，6212pic
https://www.iuii.ua.es/datasets/masati/
*）MAFAT Challenge - Fine-Grained Classification of Objects from Aerial Imagery
https://competitions.codalab.org/competitions/19854
实例分割
1）iSAID
Num：15classes，2806pic，655451instances
https://captain-whu.github.io/iSAID/
语义分割-多分类
1）GID Dataset
Size：6800*7200；6800*7200
Num：5classes，150pic；15classes，10pic
http://captain.whu.edu.cn/repository.html
https://pan.baidu.com/s/1c1lP-rlNjbfCPeaKPFteuA
链接：https://pan.baidu.com/s/1SprxRLxs4kx_s7bnSI7jNw 提取码：6xjt
2）DLRSD
Size：256*256；
Num：21classes，2100pic
基于UC Merced Land-Use Data Set进行标注
https://sites.google.com/view/zhouwx/dataset#h.p_hQS2jYeaFpV0
https://drive.google.com/file/d/1DtKiauowCB0ykjFe8v0OVvT76rEfOk0v/view
链接：https://pan.baidu.com/s/1unm9jrdmaZKWKEH0wxK7bg 提取码：f5lj
3）ISPRS 2D Semantic Labeling Contest
Vaihingen dataset
Num：6classes，33pic
http://www2.isprs.org/commissions/comm3/wg4/2d-sem-label-vaihingen.html
Postdam dataset
Num：6classes，38pic
http://www2.isprs.org/commissions/comm3/wg4/2d-sem-label-potsdam.html
需要在http://www2.isprs.org/commissions/comm3/wg4/2d-sem-label-potsdam.html中按要求发送邮件后获取下载连接
4）DeepGlobe Land Cover Classification Challenge
Size：2448*2448
Num：7classes, 803pic
https://competitions.codalab.org/competitions/18468#participate-get_starting_kit
需要在https://competitions.codalab.org/forums/15208/注册提交申请，获取下载权限
他人分享链接：
- https://github.com/zlkanata/DeepGlobe-Road-Extraction-Challenge/issues/1
- Road：https://pan.baidu.com/s/1S4fn01psiXN9mQ2CI1ZrHg password: 0qor
- land cover classification：https://pan.baidu.com/s/1KwD5M3C84jPfyCCq03tGTw password: thng
5)Zurich Summer Dataset
Num：9classes, 20pic
https://sites.google.com/site/michelevolpiresearch/data/zurich-dataset
*)SEN12MS
Num：3classes, 180748pic
https://mediatum.ub.tum.de/1474000
*)Slovenia 2017 Land Cover Classification Dataset
Size：000 * 5000
Num：940 EOPatches
http://eo-learn.sentinel-hub.com/
EvLab-SS Dataset
Size：4500*4500
Num：11classes, 35pic
http://earthvisionlab.whu.edu.cn/zm/SemanticSegmentation/index.html
*)Aeroscapes
Size：720*720
Num：11classes，3269pic
https://github.com/ishann/aeroscapes
二分类
1）Road and Building Detection Datasets
Size: 1500*1500
Num: 137+10+4=151pic
https://www.cs.toronto.edu/
2）INRIA aerial image dataset
Size：5000*5000
class：build
https://project.inria.fr/aerialimagelabeling/
https://hyper.ai/datasets/5428
3）38-Cloud: A Cloud Segmentation Dataset
Size：384*384
class：cloud，(8400 +9201)pic
https://github.com/SorourMo/38-Cloud-A-Cloud-Segmentation-Dataset
4）WHU Building Dataset
1.Aerial imagery dataset
Size：512*512
class：build，8189 pic，187000 buildings
https://study.rsgis.whu.edu.cn/pages/download/3.%20The%20cropped%20aerial%20image%20tiles%20and%20raster%20labels.zip
https://study.rsgis.whu.edu.cn/pages/download/building_dataset.html
2.Satellite dataset I (global cities)
Size：512*512
class：build，204 pic
https://study.rsgis.whu.edu.cn/pages/download/Satellite%20dataset%20Ⅰ%20(global%20cities).zip
https://study.rsgis.whu.edu.cn/pages/download/building_dataset.html
3. Satellite dataset Ⅱ (East Asia)
Size：512*512
class：build， 17388pic， 34085 buildings
https://study.rsgis.whu.edu.cn/pages/download/Satellite%20dataset%20Ⅱ%20(East%20Asia).zip
https://study.rsgis.whu.edu.cn/pages/download/building_dataset.html
*）Massachusetts Road and Building Detection Datasets
Massachusetts Roads Dataset
https://www.cs.toronto.edu/~vmnih/data/
Massachusetts Buildings Dataset
Size：1500*1500
class：build，(137+10+4)pic
https://www.cs.toronto.edu/~vmnih/data/
*)Aerial Image Segmentation Dataset
class：(building, road, and background)
https://zenodo.org/record/1154821#.XvAcO_kzaUl
变化检测
1）MtS-WH
Size：7200*6000
Num：1pic* 2time
http://sigma.whu.edu.cn/newspage.php?q=2019_03_26
2）SZTAKI AirChange Benchmark Set
Size：952*640
Num：13pic *2time
http://web.eee.sztaki.hu/remotesensing/airchange_benchmark.html 登录名：sztaki，密码：deva
3）OSCD
Num：14pic * 2time
https://rcdaudt.github.io/oscd/
4）Change Detection Dataset
size：984*740；600*500；390*200
Num：3pic* 2time
https://gitlab.citius.usc.es/hiperespectral/ChangeDetectionDataset
5） River Dataset
size：463*241
Num：1pic * 2time
https://arxiv.org/abs/1905.01662v1
https://share.weiyun.com/5ugrczK
6）WHU ： Building change detection dataset
class：build， 12796——>16077 buildings
https://study.rsgis.whu.edu.cn/pages/download/Building%20change%20detection%20dataset.zip
https://study.rsgis.whu.edu.cn/pages/download/building_dataset.html
7）Season-varying Change Detection Dataset
size：4715*2700
Num：7pic*2time
https://drive.google.com/file/d/1GX656JqqOyBi_Ef0w65kDGVto-nHrNs9/edit
8）Onera Satellite Change Detection
Num：14pic*2time
https://rcdaudt.github.io/oscd/
https://partage.mines-telecom.fr/index.php/s/G93tRIAgLs1sVBM
https://partage.mines-telecom.fr/index.php/s/2D6n03k58ygBSpu
*）AIST Building Change Detection
Size：160*160；128*128
Num：(8506+8444)pic * 2time
https://blog.csdn.net/dangdangliu/article/details/104323166
https://github.com/gistairc/ABCDdataset
*）Urban Atlas Change 2012-2018
https://land.copernicus.eu/local/urban-atlas/urban-atlas-change-2012-2018
*）High Resolution Change Detection
http://www.pshrcd.com/#/intro
目标跟踪
1）Dataset-UAV-123
https://hyper.ai/datasets/5154
https://hyper.ai/tracker/magnet?torrent=7087
图像标题 / Image caption
1）RSICD
Size：224*224
Num：10921pic，5sentences/pic
https://github.com/201528014227051/RSICD_optimal
http://pan.baidu.com/s/1bp71tE3
https://drive.google.com/open?id=0B1jt7lJDEXy3aE90cG9YSl9ScUk
2）UCM Captions
Num：2100pic，5sentences/pic
https://pan.baidu.com/s/1mjPToHq
https://github.com/201528014227051/RSICD_optimal
3）Sydney Captions
Num：613pic，5sentences/pic
https://pan.baidu.com/s/1hujEmcG
https://github.com/201528014227051/RSICD_optimal
其他
SpaceNet Challenge
包括：
https://spacenetchallenge.github.io/datasets/datasetHomePage.html
需要使用AMS公共数据集访问下载
扫一扫
1万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
