Near Real-Time Flood Mapping with Weakly Supervised Machine Learning

被引:13
作者
Vongkusolkit, Jirapa [1 ]
Peng, Bo [1 ,2 ]
Wu, Meiliu [1 ]
Huang, Qunying [1 ]
Andresen, Christian G. [3 ]
机构
[1] Univ Wisconsin, Dept Geog, Spatial Comp & Data Min Lab, Madison, WI 53706 USA
[2] PAII Inc, Palo Alto, CA 94306 USA
[3] Univ Wisconsin, Dept Geog, Madison, WI 53706 USA
基金
美国国家科学基金会; 美国食品与农业研究所;
关键词
flood extent mapping; flood detection; weakly supervised; deep learning; remote sensing; image processing; machine learning; GeoAI; natural hazards; SUPPORT VECTOR MACHINE; WATER INDEX NDWI; RANDOM FOREST; LANDSAT-TM; DELINEATION; AREAS;
D O I
10.3390/rs15133263
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Advances in deep learning and computer vision are making significant contributions to flood mapping, particularly when integrated with remotely sensed data. Although existing supervised methods, especially deep convolutional neural networks, have proved to be effective, they require intensive manual labeling of flooded pixels to train a multi-layer deep neural network that learns abstract semantic features of the input data. This research introduces a novel weakly supervised approach for pixel-wise flood mapping by leveraging multi-temporal remote sensing imagery and image processing techniques (e.g., Normalized Difference Water Index and edge detection) to create weakly labeled data. Using these weakly labeled data, a bi-temporal U-Net model is then proposed and trained for flood detection without the need for time-consuming and labor-intensive human annotations. Using floods from Hurricanes Florence and Harvey as case studies, we evaluated the performance of the proposed bi-temporal U-Net model and baseline models, such as decision tree, random forest, gradient boost, and adaptive boosting classifiers. To assess the effectiveness of our approach, we conducted a comprehensive assessment that (1) covered multiple test sites with varying degrees of urbanization, and (2) utilized both bi-temporal (i.e., pre- and post-flood) and uni-temporal (i.e., only post-flood) input. The experimental results showed that the proposed framework of weakly labeled data generation and the bi-temporal U-Net could produce near real-time urban flood maps with consistently high precision, recall, f1 score, IoU score, and overall accuracy compared with baseline machine learning algorithms.
引用
收藏
页数:25
相关论文
共 48 条
[1]   Semi-automated procedures for shoreline extraction using single RADARSAT-1 SAR image [J].
Al Fugura, A'kif ;
Billa, Lawal ;
Pradhan, Biswajeet .
ESTUARINE COASTAL AND SHELF SCIENCE, 2011, 95 (04) :395-400
[2]   Random forest in remote sensing: A review of applications and future directions [J].
Belgiu, Mariana ;
Dragut, Lucian .
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2016, 114 :24-31
[3]   SPATIOTEMPORAL CONTRASTIVE REPRESENTATION LEARNING FOR BUILDING DAMAGE CLASSIFICATION [J].
Peng, Bo ;
Huang, Qunying ;
Rao, Jinmeng .
2021 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM IGARSS, 2021, :8562-8565
[4]   Sen1Floods11: a georeferenced dataset to train and test deep learning flood algorithms for Sentinel-1 [J].
Bonafilia, Derrick ;
Tellman, Beth ;
Anderson, Tyler ;
Issenberg, Erica .
2020 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS (CVPRW 2020), 2020, :835-845
[5]  
Chen T., 2020, P INT C MACH LEARN P
[6]   Image Segmentation using K-means Clustering Algorithm and Subtractive Clustering Algorithm [J].
Dhanachandra, Nameirakpam ;
Manglem, Khumanthem ;
Chanu, Yambem Jina .
ELEVENTH INTERNATIONAL CONFERENCE ON COMMUNICATION NETWORKS, ICCN 2015/INDIA ELEVENTH INTERNATIONAL CONFERENCE ON DATA MINING AND WAREHOUSING, ICDMW 2015/NDIA ELEVENTH INTERNATIONAL CONFERENCE ON IMAGE AND SIGNAL PROCESSING, ICISP 2015, 2015, 54 :764-771
[7]   Urban Flood Mapping Based on Unmanned Aerial Vehicle Remote Sensing and Random Forest Classifier-A Case of Yuyao, China [J].
Feng, Quanlong ;
Liu, Jiantao ;
Gong, Jianhua .
WATER, 2015, 7 (04) :1437-1455
[8]   Automated Water Extraction Index: A new technique for surface water mapping using Landsat imagery [J].
Feyisa, Gudina L. ;
Meilby, Henrik ;
Fensholt, Rasmus ;
Proud, Simon R. .
REMOTE SENSING OF ENVIRONMENT, 2014, 140 :23-35
[9]   Deep Convolutional Neural Network for Flood Extent Mapping Using Unmanned Aerial Vehicles Data [J].
Gebrehiwot, Asmamaw ;
Hashemi-Beni, Leila ;
Thompson, Gary ;
Kordjamshidi, Parisa ;
Langan, Thomas E. .
SENSORS, 2019, 19 (07)
[10]   Semantic Segmentation and Edge Detection-Approach to Road Detection in Very High Resolution Satellite Images [J].
Ghandorh, Hamza ;
Boulila, Wadii ;
Masood, Sharjeel ;
Koubaa, Anis ;
Ahmed, Fawad ;
Ahmad, Jawad .
REMOTE SENSING, 2022, 14 (03)