Scale Sequence Joint Deep Learning (SS-JDL) for land use and land cover classification

doi:10.1016/j.rse.2019.111593

	Scale Sequence Joint Deep Learning (SS-JDL) for land use and land cover classification
	Zhang, Ce 4,9; Harrison, Paula A.4; Pan, Xin 1,5; Li, Huapeng 6; Sargent, Isabel 7; Atkinson, Peter M.2,3,8,10
刊名	REMOTE SENSING OF ENVIRONMENT
	2020-02-01
卷号	237 页码:16
关键词	Multi-scale deep learning Optimal scale selection Convolutional neural network Joint classification Hierarchical representations
ISSN号	0034-4257
DOI	10.1016/j.rse.2019.111593
通讯作者	Zhang, Ce(c.zhang9@lancaster.ac.uk) ; Atkinson, Peter M.(pma@lancaster.ac.uk)
英文摘要	Choosing appropriate scales for remotely sensed image classification is extremely important yet still an open question in relation to deep convolutional neural networks (CNN), due to the impact of spatial scale (i.e., input patch size) on the recognition of ground objects. Currently, the optimal scale selection processes are extremely cumbersome and time-consuming requiring repetitive experiments involving trial-and-error procedures, which significantly reduce the practical utility of the corresponding classification methods. This issue is crucial when trying to classify large-scale land use (LU) and land cover (LC) jointly (Zhang et al., 2019). In this paper, a simple and parsimonious Scale Sequence Joint Deep Learning (SS-JDL) method is proposed for joint LU and LC classification, in which a sequence of scales is embedded in the iterative process of fitting the joint distribution implicit in the joint deep learning (JDL) method, thus, replacing the previous paradigm of scale selection. The sequence of scales, derived autonomously and used to define the CNN input patch sizes, provides consecutive information transmission from small-scale features to large-scale representations, and from simple LC states to complex LU characterisations. The effectiveness of the novel SS-JDL method was tested on aerial digital photography of three complex and heterogeneous landscapes, two in Southern England (Bournemouth and Southampton) and one in North West England (Manchester). Benchmark comparisons were provided in the form of a range of LU and LC methods, including the state-of-the-art joint deep learning (JDL) method. The experimental results demonstrated that the SS-JDL consistently outperformed all of the state-of-the-art baselines in terms of both LU and LC classification accuracies, as well as computational efficiency. The proposed SS-JDL method, therefore, represents a fast and effective implementation of the state-of-the-art JDL method. By creating a single, unifying joint distribution framework for classifying higher order feature representations, including LU, the SS-JDL method has the potential to transform the classification paradigm in remote sensing, and in machine learning more generally.
资助项目	Centre of Excellence in Environmental Data Science (CEEDS) - Lancaster University ; Centre of Excellence in Environmental Data Science (CEEDS) - UK Centre for Ecology Hydrology ; National Key Research and Development Program of China[2016YFB0502300] ; National Natural Science Foundation of China[41871236]
WOS关键词	NEURAL-NETWORKS ; SCENE CLASSIFICATION ; TEXTURE ; GEOBIA ; CNN
WOS研究方向	Environmental Sciences & Ecology ; Remote Sensing ; Imaging Science & Photographic Technology
语种	英语
出版者	ELSEVIER SCIENCE INC
WOS记录号	WOS:000509819300041
资助机构	Centre of Excellence in Environmental Data Science (CEEDS) - Lancaster University ; Centre of Excellence in Environmental Data Science (CEEDS) - UK Centre for Ecology Hydrology ; National Key Research and Development Program of China ; National Natural Science Foundation of China
内容类型	期刊论文
源URL	[http://ir.igsnrr.ac.cn/handle/311030/132201]
专题	中国科学院地理科学与资源研究所
通讯作者	Zhang, Ce; Atkinson, Peter M.
作者单位	1.Changchun Inst Technol, Sch Comp Technol & Engn, Changchun 130012, Peoples R China 2.Univ Southampton, Geog & Environm Sci, Southampton SO17 1BJ, Hants, England 3.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, 11A Datun Rd, Beijing 100101, Peoples R China 4.Ctr Ecol & Hydrol, Lib Ave, Lancaster LA1 4AP, England 5.Changchun Inst Technol, Key Lab Changbai Mt Hist Culture & VR Technol Rec, Changchun 130012, Peoples R China 6.Chinese Acad Sci, Northeast Inst Geog & Agroecol, Changchun 130102, Peoples R China 7.Ordnance Survey, Adanac Dr, Southampton SO16 0AS, Hants, England 8.Univ Lancaster, Fac Sci & Technol, Lancaster LA1 4YR, England 9.Univ Lancaster, Lancaster Environm Ctr, Lancaster LA1 4YQ, England 10.Queens Univ Belfast, Sch Nat & Built Environm, Belfast BT7 1NN, Antrim, North Ireland
推荐引用方式 GB/T 7714	Zhang, Ce,Harrison, Paula A.,Pan, Xin,et al. Scale Sequence Joint Deep Learning (SS-JDL) for land use and land cover classification[J]. REMOTE SENSING OF ENVIRONMENT,2020,237:16.
APA	Zhang, Ce,Harrison, Paula A.,Pan, Xin,Li, Huapeng,Sargent, Isabel,&Atkinson, Peter M..(2020).Scale Sequence Joint Deep Learning (SS-JDL) for land use and land cover classification.REMOTE SENSING OF ENVIRONMENT,237,16.
MLA	Zhang, Ce,et al."Scale Sequence Joint Deep Learning (SS-JDL) for land use and land cover classification".REMOTE SENSING OF ENVIRONMENT 237(2020):16.