基于深度异构迁移学习的水稻遥感影像提取

doi:10.14088/j.cnki.issn0439-8114.2024.08.039

湖北农业科学 ›› 2024, Vol. 63 ›› Issue (8): 236-242.doi: 10.14088/j.cnki.issn0439-8114.2024.08.039

基于深度异构迁移学习的水稻遥感影像提取

邱儒琼^1,2, 何丽华³, 李孟璠²

1.中国地质大学（武汉）国家地理信息系统工程技术研究中心,武汉 430074;
2.湖北省发展规划研究院有限公司,武汉 430071;
3.湖北省地理国情监测中心,武汉 430071

收稿日期:2022-12-12 出版日期:2024-08-25 发布日期:2024-09-05
作者简介:邱儒琼（1977-）,女,湖北公安人,正高级工程师,在读博士研究生,主要从事遥感数据智能处理的理论和应用研究,（电话）13871363256（电子信箱）404251392@qq.com。
基金资助:
湖北省自然资源厅科研项目（ZRZY2021KJ03）

Extraction of rice from remote sensing images based on deep heterogeneous transfer learning

QIU Ru-qiong^1,2, HE Li-hua³, LI Meng-fan²

1. National Engineering Research Center of Geographic Information System, China University of Geosciences（Wuhan）, Wuhan 430074, China;
2. Hubei Development Planning Research Institute Co., Ltd., Wuhan 430071, China;
3. Hubei Geography National Condition Monitoring Center, Wuhan 430071, China

Received:2022-12-12 Published:2024-08-25 Online:2024-09-05

摘要/Abstract

摘要： 为了实现在目标域仅有无标注样本的条件下对异构遥感影像上的水稻提取模型进行高质量构建和复用,构建了一种基于时空约束的深度异构特征迁移学习模型。首先,基于空间位置构建源域和目标域无标签样本组,并提取其深度特征;其次,构建异构特征迁移模型,创建同名样本特征转换、同名样本特征正则、样本重建损失函数,减少特征负迁移影响,实现异构特征的精准迁移;最后,建立半监督分类模型,通过引入HingLoss损失来消除错误伪标签的影响,实现分类精度的提高。结果表明,本研究方法能够实现不同分辨率下影像间的样本特征迁移,相较于未经过特征迁移的情况,准确率提升了27.68个百分点,F1分数提升了17.3个百分点。

关键词: 无标注样本, 水稻提取, 高分辨率遥感影像, 深度异构迁移学习

Abstract: In order to achieve high-quality construction and reuse of rice extraction models from remote sensing images based on heterogeneous with only unlabeled samples in the target domain, a deep heterogeneous feature transfer learning model based on temporal and spatial constraints was constructed. Firstly, unlabeled sample groups in the source domains and target domains were constructed based on spatial location, and their deep features were extracted; secondly, in order to reduce the negative transfer impact of features and realize precise transfer of heterogeneous features, a heterogeneous feature transferring model was constructed by using a composite loss function including corresponding sample feature conversion loss, corresponding sample feature regular loss, and sample reconstruction loss; finally, in order to improve the accuracy of classification, a semi-supervised classification model was established, and HingLoss was introduced to eliminate the impact of wrong pseudo labels. The results showed that the research method could realize sample feature transfer between images at different resolutions. Compared with the case without feature transfer, the accuracy rate was improved by 27.68 percentage points, and the F1 score was improved by 17.3 percentage points.

Key words: unlabeled sample, extraction of rice, high resolution remote sensing images, deep heterogeneous transfer learning

中图分类号:

P237

邱儒琼, 何丽华, 李孟璠. 基于深度异构迁移学习的水稻遥感影像提取[J]. 湖北农业科学, 2024, 63(8): 236-242.

QIU Ru-qiong, HE Li-hua, LI Meng-fan. Extraction of rice from remote sensing images based on deep heterogeneous transfer learning[J]. HUBEI AGRICULTURAL SCIENCES, 2024, 63(8): 236-242.

参考文献

[1] 谢登峰,张锦水,潘耀忠,等.Landsat 8和MODIS融合构建高时空分辨率数据识别秋粮作物[J].遥感学报,2015,19(5):791-805.
[2] 陈安旭,李月臣.基于Sentinel-2影像的西南山区不同生长期水稻识别[J].农业工程学报,2020,36(7):192-199.
[3] 高永康,王腊红,陈家赢,等.一种利用形态相似性的水稻信息提取算法[J].遥感信息,2020,35(2):76-86.
[4] 许童羽,胡开越,周云成,等.基于CART决策树和BP神经网络的landsat 8影像粳稻提取方法[J].沈阳农业大学学报,2020, 51(2):169-176.
[5] 王建勋,华丽,邓世超,等.基于GF-1与MODIS时空融合的南方丘陵区水稻提取研究[J].中国农业资源与区划,2019,40(5):37-46.
[6] 赵亚杰,黄进良,王立辉,等.基于时空融合NDVI及物候特征的江汉平原水稻种植区提取研究[J].长江流域资源与环境,2020,29(2):424-432.
[7] 赖格英,杨星卫.南方丘陵地区水稻种植面积遥感信息提取的试验[J].应用气象学报,2000,11(1):47-56.
[8] 李石,宋晓巍,张菁,等.基于遥感数据的沈阳水稻种植面积提取及精度分析[J].大麦与谷类科学,2016,33(4):63-66.
[9] KUSSUL N,LAVRENIUK M,SKAKUN S,et al.Deep learning classification of land cover and crop types using remote sensing data[J].IEEE geoscience and remote sensing letters,2017,14(5):778-782.
[10] WEI S,ZHANG H,WANG C,et al.Multi-temporal SAR data large-scale crop mapping based on U-Net model[J].Remote sensing,2019,11(1):68.
[11] YANG X,YE Y,LI X,et al.Hyperspectral image classification with deep learning models[J].IEEE transactions on geoscience and remote sensing,2018,56(99):5408-5423.
[12] ZHONG L,HU L,ZHOU H.Deep learning based multi-temporal crop classification[J].Remote sensing of environment,2019,221:430-443.
[13] HUETE A,DIDAN K,MIURA T,et al.Overview of the radiometric and biophysical performance of the MODIS vegetation indices[J].Remote sensing of environment,2002,83(1-2):195-213.
[14] BLITZER J C,MCDONALD R T,PEREIRA F C N. Domain adaptation with structural correspondence learning[A].Proceedings of conference on empirical methods in natural language processing[C].Sydney,Australia,2006.120-128.
[15] PRETTENHOFER P,STEIN B B.Cross-language text classification using structural correspondence learning[A].Proceedings of 48th annual meeting of the association for computational linguistics[C]. Uppsala,Sweden,2010.1118-1127.
[16] NAM J,FU W,KIM S,et al.Heterogeneous defect prediction[J].IEEE transactions on software engineering,2017,44(9):874-896.
[17] ZHOU J T,PAN S J,TSANG I W,et al.Hybrid heterogeneous transfer learning through deep learning[A]. Proceedings of AAAI conference on artificial intelligence[C].Québec,Canada,2014.2213-2219.
[18] 赵亮,吴立宗,彭红春,等.基于迁移学习的高分辨率遥感影像道路信息提取[J].江苏海洋大学学报(自然科学版),2020,29(2):69-72.
[19] 周绍光,吴昊,赵婵娟,等.利用同质区特性的高光谱图像迁移学习分类[J].计算机工程与应用,2021,57(21):224-233.
[20] MEI S,JI J,HOU J,et al.Learning sensor-specific spatial-spectral features of hyperspectral images via convolutional neural networks[J].IEEE transactions on geoscience and remote sensing,2017,55(8):4520-4533.
[21] WANG Z,DU B,SHI Q,et al.Domain adaptation with discriminative distribution and manifold embedding for hyperspectral image classification[J].IEEE transactions on geoscience and remote sensing,2019,16(7):1155-1159.
[22] 施慧慧,徐雁南,滕文秀,等.高分辨率遥感影像深度迁移可变形卷积的场景分类法[J].测绘学报,2021,50(5):652-663.
[23] 张帆. 面向高分辨率遥感影像分析的深度学习方法研究[D].武汉:武汉大学,2017.
[24] 周瑜. 基于卷积神经网络的高分辨率遥感图像分类方法研究[D].重庆:重庆邮电大学,2018.
[25] 王海军. 深度卷积神经网络在遥感影像分类的应用研究[D].北京:中国地质大学(北京),2018.
[26] 许慧敏. 基于深度学习U-Net模型的高分辨率遥感影像分类方法研究[D].成都:西南交通大学,2018.
[27] 闫苗,赵红东,李宇海,等.基于卷积神经网络的高光谱遥感地物多分类识别[J].激光与光电子学进展,2019,56(2):191-198.
[28] KRIZHEVSKY A,SUTSKEVER I,HINTON G E.ImageNet classification with deep convolutional neural networks[J].Communications of the ACM, 2017,60(6):84-90.

基于深度异构迁移学习的水稻遥感影像提取

Extraction of rice from remote sensing images based on deep heterogeneous transfer learning

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 3

编辑推荐

Metrics

本文评价

[1]	邱儒琼, 彭少坤, 李孟璠. 基于深度学习和遥感数据的水稻种植面积提取[J]. 湖北农业科学, 2023, 62(11): 176-182.
[2]	马海荣, 沈祥成, 罗治情, 陈聘婷, 官波. 基于建筑遥感指数与面向对象的高分影像农村建筑物信息提取[J]. 湖北农业科学, 2021, 60(23): 148-152.
[3]	王克晓, 周蕊, 虞豹, 黄祥, 王茜. 基于随机森林的水稻信息提取研究[J]. 湖北农业科学, 2018, 57(21): 143-146.