Breaking the resolution barrier: A low-to-high network for large-scale high-resolution land-cover mapping using low-resolution labels

被引:59
作者
Li, Zhuohong [1 ]
Zhang, Hongyan [1 ]
Lu, Fangxiao [1 ]
Xue, Ruoyao [2 ]
Yang, Guangyi [2 ]
Zhang, Liangpei [1 ]
机构
[1] Wuhan Univ, State Key Lab Informat Engn Surveying Mapping & R, Wuhan 430079, Peoples R China
[2] Wuhan Univ, Sch Elect Informat, Wuhan 430079, Peoples R China
基金
中国国家自然科学基金;
关键词
Multi-resolution; Land-cover mapping; Semantic segmentation; Low-to-high task; SEMANTIC SEGMENTATION; RANDOM FOREST; CLASSIFICATION; SELECTION; MAPS;
D O I
10.1016/j.isprsjprs.2022.08.008
中图分类号
P9 [自然地理学];
学科分类号
0705 ; 070501 ;
摘要
Large-scale high-resolution land-cover mapping is a way to comprehend the Earth's surface and resolve the ecological and resource challenges facing humanity. High-resolution (<= 1 m) remotely sensed images can now be captured more easily, with wider coverage, as sensors and satellites develop. Nevertheless, the synchronous renewal of land-cover maps is still challenging when using the common land-cover mapping methods, due to the requirement for high-resolution land-cover labels. Abundant low-resolution (similar to 30 m) land-cover products are available for use as alternative label sources, but the resolution gap between these products and the growing volume of high-resolution imagery is a barrier yet to be overcome. In this paper, to break through this obstacle, we propose a low-to-high network (L2HNet) to automatically generate high-resolution land-cover maps from high-resolution images by taking only low-resolution land-cover products as the training labels, thus getting rid of the requirement for finely labeled samples during the large-scale map updating process. Firstly, to obtain the mapping results with rich details, we propose a resolution-preserving (RP) backbone that contains parallel multi-scale convolutional layers for extracting the high-resolution features from the images. Furthermore, to settle the label noise issue caused by the mismatched resolution, a confident area selection (CAS) module and a low-to-high (L2H) loss function, with weak and unsupervised strategies, are designed for obtaining reliable supervision information from the coarse labels. The experimental results obtained for six administrative states located in the Chesapeake Bay watershed of the United States show that L2HNet outperforms several of the state-of-the-art methods and the mainstream land-cover mapping methods in creating 1-m resolution land-cover maps by taking 30-m resolution land-cover products as training labels. As a further application, L2HNet was also adopted to produce the first 1-m resolution land-cover map with level II classification hierarchy for the entire state of Maryland in the United States, which covers an area of about 33,872 km(2). The land-cover map of Maryland is publicly available at http://hipag.whu.edu.cn/L2HNet.html.
引用
收藏
页码:244 / 267
页数:24
相关论文
共 65 条
[1]   LandCover.ai: Dataset for Automatic Mapping of Buildings, Woodlands, Water and Roads from Aerial Imagery [J].
Boguszewski, Adrian ;
Batorski, Dominik ;
Ziemba-Jankowska, Natalia ;
Dziedzic, Tomasz ;
Zambrzycka, Anna .
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS, CVPRW 2021, 2021, :1102-1110
[2]   Toward the Automatic Updating of Land-Cover Maps by a Domain-Adaptation SVM Classifier and a Circular Validation Strategy [J].
Bruzzone, Lorenzo ;
Marconcini, Mattia .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2009, 47 (04) :1108-1122
[3]   Evaluation of Random Forest and Adaboost tree-based ensemble classification and spectral band selection for ecotope mapping using airborne hyperspectral imagery [J].
Chan, Jonathan Cheung-Wai ;
Paelinckx, Desire .
REMOTE SENSING OF ENVIRONMENT, 2008, 112 (06) :2999-3011
[4]  
Chaurasia A, 2017, 2017 IEEE VISUAL COMMUNICATIONS AND IMAGE PROCESSING (VCIP)
[5]   Urban road mapping based on an end-to-end road vectorization mapping network framework [J].
Chen, Dingyuan ;
Zhong, Yanfei ;
Zheng, Zhuo ;
Ma, Ailong ;
Lu, Xiaoyan .
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2021, 178 :345-365
[6]   Global land cover mapping at 30 m resolution: A POK-based operational approach [J].
Chen, Jun ;
Chen, Jin ;
Liao, Anping ;
Cao, Xin ;
Chen, Lijun ;
Chen, Xuehong ;
He, Chaoying ;
Han, Gang ;
Peng, Shu ;
Lu, Miao ;
Zhang, Weiwei ;
Tong, Xiaohua ;
Mills, Jon .
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2015, 103 :7-27
[7]   Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation [J].
Chen, Liang-Chieh ;
Zhu, Yukun ;
Papandreou, George ;
Schroff, Florian ;
Adam, Hartwig .
COMPUTER VISION - ECCV 2018, PT VII, 2018, 11211 :833-851
[8]   Agriculture-Vision: A Large Aerial Image Database for Agricultural Pattern Analysis [J].
Chiu, Mang Tik ;
Xu, Xingqian ;
Wei, Yunchao ;
Huang, Zilong ;
Schwing, Alexander G. ;
Brunner, Robert ;
Khachatrian, Hrant ;
Karapetyan, Hovnatan ;
Dozier, Ivan ;
Rose, Greg ;
Wilson, David ;
Tudor, Adrian ;
Hovakimyan, Naira ;
Huang, Thomas S. ;
Shi, Honghui .
2020 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2020, :2825-2835
[9]   Land cover classification with coarse spatial resolution data to derive continuous and discrete maps for complex regions [J].
Colditz, R. R. ;
Schmidt, M. ;
Conrad, C. ;
Hansen, M. C. ;
Dech, S. .
REMOTE SENSING OF ENVIRONMENT, 2011, 115 (12) :3264-3275
[10]   ResUNet-a: A deep learning framework for semantic segmentation of remotely sensed data [J].
Diakogiannis, Foivos, I ;
Waldner, Francois ;
Caccetta, Peter ;
Wu, Chen .
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2020, 162 :94-114