Automatic land cover classification of geo-tagged field photos by deep learning

doi:10.1016/j.envsoft.2017.02.004

	Automatic land cover classification of geo-tagged field photos by deep learning
	Xu, Guang 1; Zhu, Xuan 1; Fu, Dongjie 2; Dong, Jinwei 3,4,5; Xiao, Xiangming 4,5
刊名	ENVIRONMENTAL MODELLING & SOFTWARE
	2017-05-01
卷号	91 页码:127-134
关键词	Deep learning Convolutional neural network Transfer learning Multinomial logistic regression Land cover Crowdsourced photos
ISSN号	1364-8152
DOI	10.1016/j.envsoft.2017.02.004
通讯作者	Xu, Guang(xg1990@gmail.com)
英文摘要	With more and more crowdsourcing geo-tagged field photos available online, they are becoming a potentially valuable source of information for environmental studies. However, the labelling and recognition of these photos are time-consuming. To utilise such information, a land cover type recognition model for field photos was proposed based on the deep learning technique. This model combines a pre-trained convolutional neural network (CNN) as the image feature extractor and the multinomial logistic regression model as the feature classifier. The pre-trained CNN model Inception-v3 was used in this study. The labelled field photos from the Global Geo-Referenced Field Photo Library (http://eomf.ou.eduiphotos) were chosen for model training and validation. The results indicated that our recognition model achieved an acceptable accuracy (48.40% for top-1 prediction and 76.24% for top-3 prediction) of land cover classification. With accurate self-assessment of confidence, the model can be applied to classify numerous online geo-tagged field photos for environmental information extraction. (C) 2017 Elsevier Ltd. All rights reserved.
资助项目	NASA Land Use and Land Cover Change program[NNX14AD78G] ; Key Research Program of Frontier Sciences ; Chinese Academy of Sciences[QYZDB-SSW-DQC005] ; Thousand Youth Talents Plan ; Youth Science Funds of State Key Laboratory of Resources and Environmental Information System[O8R8A080YA] ; National Science Foundation of China[41501473] ; Institute of Geographic Sciences and Natural Resources Research[Y6V60206YZ]
WOS关键词	DECIDUOUS RUBBER PLANTATIONS ; FACE DETECTION ; IMAGERY ; BASIN ; RECOGNITION ; PALSAR
WOS研究方向	Computer Science ; Engineering ; Environmental Sciences & Ecology
语种	英语
出版者	ELSEVIER SCI LTD
WOS记录号	WOS:000399868000009
资助机构	NASA Land Use and Land Cover Change program ; Key Research Program of Frontier Sciences ; Chinese Academy of Sciences ; Thousand Youth Talents Plan ; Youth Science Funds of State Key Laboratory of Resources and Environmental Information System ; National Science Foundation of China ; Institute of Geographic Sciences and Natural Resources Research
内容类型	期刊论文
源URL	[http://ir.igsnrr.ac.cn/handle/311030/62620]
专题	中国科学院地理科学与资源研究所
通讯作者	Xu, Guang
作者单位	1.Monash Univ, Sch Earth Atmosphere & Environm, Clayton Campus, Clayton, Vic 3800, Australia 2.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, State Key Lab Resources & Environm Informat Syst, Beijing 100101, Peoples R China 3.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, Key Lab Land Surface Pattern & Simulat, Beijing 100101, Peoples R China 4.Univ Oklahoma, Dept Microbiol & Plant Biol, Norman, OK 73019 USA 5.Univ Oklahoma, Ctr Spatial Anal, Norman, OK 73019 USA
推荐引用方式 GB/T 7714	Xu, Guang,Zhu, Xuan,Fu, Dongjie,et al. Automatic land cover classification of geo-tagged field photos by deep learning[J]. ENVIRONMENTAL MODELLING & SOFTWARE,2017,91:127-134.
APA	Xu, Guang,Zhu, Xuan,Fu, Dongjie,Dong, Jinwei,&Xiao, Xiangming.(2017).Automatic land cover classification of geo-tagged field photos by deep learning.ENVIRONMENTAL MODELLING & SOFTWARE,91,127-134.
MLA	Xu, Guang,et al."Automatic land cover classification of geo-tagged field photos by deep learning".ENVIRONMENTAL MODELLING & SOFTWARE 91(2017):127-134.