A new multi-source remote sensing image sample dataset with high resolution for flood area extraction:GF-FloodNet

作     者：Yuwei Zhang Peng Liu Lajiao Chen Mengzhen Xu Xingyan Guo Lingjun Zhao 

作者机构：Aerospace Information Research InstituteChinese Academy of SciencesBeijingPeople’s Republic of China School of ElectronicElectrical and Communication EngineeringUniversity of Chinese Academy of SciencesBeijingPeople’s Republic of China State Key Laboratory of Hydroscience and EngineeringDepartment of Hydraulic EngineeringTsinghua UniversityBeijingPeople’s Republic of China 

出 版 物：《International Journal of Digital Earth》 (国际数字地球学报（英文）)

年 卷 期：2023年第16卷第1期

页      面：2522-2554页

核心收录：

学科分类：0810[工学-信息与通信工程] 08[工学] 081002[工学-信号与信息处理] 

基　　金：supported by the National Natural Science Foundation of China under Grant number U2243222 42071413 and 41971397 

主　　题：Flood area extraction dataset construction multi-source remote sensing data deep learning 

摘      要：Deep learning algorithms show good prospects for remote sensingflood *** mostly rely on huge amounts of labeled ***,there is a lack of available labeled data in actual *** this paper,we propose a high-resolution multi-source remote sensing dataset forflood area extraction:***-FloodNet contains 13388 samples from Gaofen-3(GF-3)and Gaofen-2(GF-2)*** use a multi-level sample selection and interactive annotation strategy based on active learning to construct *** with otherflood-related datasets,GF-FloodNet not only has a spatial resolution of up to 1.5 m and provides pixel-level labels,but also consists of multi-source remote sensing *** thoroughly validate and evaluate the dataset using several deep learning models,including quantitative analysis,qualitative analysis,and validation on large-scale remote sensing data in real *** results reveal that GF-FloodNet has significant advantages by multi-source *** can support different deep learning models for training to extractflood *** should be a potential optimal boundary for model training in any deep learning *** boundary seems close to 4824 samples in *** provide GF-FloodNet at https://***/datasets/pengliuair/gf-floodnet and https://***/s/1vdUCGNAfFwG5UjZ9RLLFMQ?pwd=8v6o.