基于深度学习融合OBIA的黄土高原小流域淤地坝系提取

doi:10.12118/j.issn.1000-6060.2023.057

干旱区地理 ›› 2023, Vol. 46 ›› Issue (11): 1803-1812.doi: 10.12118/j.issn.1000-6060.2023.057

基于深度学习融合OBIA的黄土高原小流域淤地坝系提取

钱伟¹(),王春^2,³(),代文⁴,卢旺达¹,李敏^3,⁴,陶宇^2,³,李梦琪⁴

1.南京信息工程大学遥感与测绘工程学院，江苏南京 211800
2.滁州学院地理信息与旅游学院，安徽滁州 239000
3.实景地理环境安徽省重点实验室，安徽滁州 239000
4.南京信息工程大学地理科学学院，江苏南京 211800

收稿日期:2023-02-14 修回日期:2023-03-30 出版日期:2023-11-25 发布日期:2023-12-05
通讯作者: 王春（1975-），男，博士，教授，主要从事实景三维建模研究. E-mail: wangchun93@126.com
作者简介:钱伟（1996-），男，硕士研究生，主要从事数字地形分析研究. E-mail: 1003435237@qq.com
基金资助:
安徽高校省级自然科学研究重大项目(KJ2021ZD0130);实景地理环境智能科技滁州市“113”产业创新团队;江苏省高等学校自然科学研究项目(22KJB170016);实景地理环境安徽省重点实验室开放课题(2022PGE013);安徽省高等学校科研计划项目(2022AH010066)

Extraction of check dam system in small watershed of Loess Plateau based on deep learning and OBIA

QIAN Wei¹(),WANG Chun^2,³(),DAI Wen⁴,LU Wangda¹,LI Min^3,⁴,TAO Yu^2,³,LI Mengqi⁴

1. School of Remote Sensing and Surveying Engineering, Nanjing University of Information Science and Technology, Nanjing 211800, Jiangsu, China
2. School of Geographic Information and Tourism, Chuzhou University, Chuzhou 239000, Anhui, China
3. Anhui Provincial Key Laboratory of Real Scene Simulation and Environmental Display, Chuzhou University, Chuzhou 239000, Anhui, China
4. School of Geography Science, Nanjing University of Information Science and Technology, Nanjing 211800, Jiangsu, China

Received:2023-02-14 Revised:2023-03-30 Online:2023-11-25 Published:2023-12-05

摘要/Abstract

摘要：

淤地坝对于防治黄土高原水土流失有不可替代的作用，因此精确提取淤地范围和淤地坝点位对研究黄土高原水土有重要意义。现有图像分类方法中缺乏对淤地坝地形特征的考虑，容易被误判为梯田或土堆。除此之外，自动提取研究多集中于淤地范围提取，淤地坝点位仍依赖人工判读。因此，提出一种自动提取淤地坝系的方法：通过深度学习融合面向对象的影像分析（OBIA）方法提取韭园沟流域淤地范围，再利用水文分析方法提取淤地坝点位。结果表明：本方法提取的淤地范围精准率、召回率、F1Score分别为81.97%、90.94%、89.70%，F1Score与仅使用OBIA方法相比提升了21.94%。淤地坝点位的自动识别准确率为81.08%，完整率为88.89%，与前人目视解译的准确度相近，并实现了淤地坝范围和淤地坝点位的全要素提取。研究结果可为黄土高原淤地坝空间布局优化和水土流失评估等分析提供重要基础数据。

关键词: 淤地范围提取, 淤地坝点位提取, 面向对象的影像分析（OBIA）, U-Net框架, 黄土高原

Abstract:

Check dams play an irreplaceable role in preventing and controlling soil erosion in the Loess Plateau of China. Accurately extracting check dam areas and their locations holds immense importance for soil and water conservation studies in the Loess Plateau of China. However, existing image classification methods lack consideration of the topographic features of check dams, leading to many misidentifications. This study proposes a method that combines object-based image analysis (OBIA) and deep learning to precisely identify check dam areas and their locations. First, a multiscale segmentation method based on an object-oriented approach is used to segment the multisource data. Deep learning is applied to identify check dams. The results from these processes were fused using majority voting to enhance the accuracy of check dam area extraction. Second, the river network within the basin is extracted based on digital elevation model data. Finally, the check dam areas and river network are intersected to determine the check dam locations. The results show that the precision and recall rates of this method are 81.97% and 90.94%, respectively, and an F1Score value of 89.70%, which is 21.94% higher than the methods using only OBIA. Additionally, the accuracy and completeness of automatically identified check dam locations are 81.08% and 88.89%, respectively. This method provides important data for the optimal spatial layout of check dams in the Loess Plateau and for the evaluation of soil and water loss.

Key words: silt range extraction, extraction of check dam points, object-bases image analysis (OBIA), U-Net framework, Loess Plateau

钱伟, 王春, 代文, 卢旺达, 李敏, 陶宇, 李梦琪. 基于深度学习融合OBIA的黄土高原小流域淤地坝系提取[J]. 干旱区地理, 2023, 46(11): 1803-1812.

QIAN Wei, WANG Chun, DAI Wen, LU Wangda, LI Min, TAO Yu, LI Mengqi. Extraction of check dam system in small watershed of Loess Plateau based on deep learning and OBIA[J]. Arid Land Geography, 2023, 46(11): 1803-1812.

图/表 9

图1

图2

图3

图4

表1

混淆矩阵"

实际类	预测类
实际类	淤地范围	非淤地范围
淤地范围	$T P$	$F N$
非淤地范围	$F P$	$T N$

表1

图5

表2

图6

图7

参考文献 26

[1]	信忠保, 蔡强国, 宁堆虎, 等. 淤地坝与check dam的差异及其英文译法[J]. 中国水土保持科学, 2022, 20(3): 102-108.
	[ Xin Zhongbao, Cai Qiangguo, Ning Duihu, et al. A new English term of ‘Yudiba dam’ and its differences from check dam[J]. Science of Soil and Water Conservation, 2022, 20(3): 102-108. ]
[2]	马建全, 吴钶桥, 李识博, 等. 黄土高原淤地坝溃决型泥石流易发性评价[J]. 科学技术与工程, 2022, 22(17): 6869-6879.
	[ Ma Jianquan, Wu Keqiao, Li Shibo, et al. Evaluation on the vulnerability of debris flow caused by the collapse of check dam on the Loess Plateau[J]. Science, Technology and Engineering, 2022, 22(17): 6869-6879. ]
[3]	孙维营, 史学建, 张攀, 等. 小流域不同淤地坝系布局拦沙级联效应研究[J]. 人民黄河, 2016, 38(9): 82-85.
	[ Sun Weiying, Shi Xuejian, Zhang Pan, et al. Study on the cascade effect of sediment retaining by different warping dam systems in small watershed[J]. Yellow River, 2016, 38(9): 82-85. ]
[4]	杨松, 张岩, 阿尼克孜·肉孜, 等. 陕北黄土区植被恢复对切沟发育速率的影响[J]. 中国水土保持科学, 2016, 14(4): 18-25.
	[ Yang Song, Zhang Yan, Rouzi Anikezi, et al. Effects of vegetation restoration on gully development rates in northern Shaanxi[J]. Science of Soil and Water Conservation, 2016, 14(4): 18-25. ]
[5]	马煜栋, 杨帅, 韩静, 等. 陕西榆林地区无定河流域淤地坝遥感解译[J]. 干旱区地理, 2022, 45(3): 786-791.
	[ Ma Yudong, Yang Shuai, Han Jing, et al. Remote sensing interpretation of check dams in Wuding River Basin in Yulin Area of Shaanxi Province[J]. Arid Land Geography, 2022, 45(3): 786-791. ]
[6]	Gong C, Han J W, Lu X Q. Remote sensing image scene classification: Benchmark and state of the art[J]. Proceedings of the IEEE, 2017, 105(10): 1865-1883. doi: 10.1109/JPROC.2017.2675998
[7]	Talukdar S, Singha P, Mahato S, et al. Land-use land-cover classification by machine learning classifiers for satellite observations: A review[J]. Remote Sensing, 2020, 12(7): 1135, doi: 10.3390/rs12071135.
[8]	Hossain M D, Chen D. Segmentation for object-based image analysis (OBIA): A review of algorithms and challenges from remote sensing perspective[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2019, 150: 115-134. doi: 10.1016/j.isprsjprs.2019.02.009
[9]	Cheng G, Han J. A survey on object detection in optical remote sensing images[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2016, 117: 11-28. doi: 10.1016/j.isprsjprs.2016.03.014
[10]	Qiang Q Y, Huan F S, Li T W, et al. Deep learning in environmental remote sensing: Achievements and challenges[J]. Remote Sensing of Environment, 2020, 241: 111716, doi: 10.1016/j.rse.2020.111716.
[11]	Cheng G, Xing X X, Jun W H, et al. Remote sensing image scene classification meets deep learning: Challenges, methods, benchmarks, and opportunities[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020, 13: 3735-3756. doi: 10.1109/JSTARS.4609443
[12]	Shakya A, Biswas M, Pal M. Parametric study of convolutional neural network based remote sensing image classification[J]. International Journal of Remote Sensing, 2021, 42(7): 2663-2685. doi: 10.1080/01431161.2020.1857877
[13]	Lecun Y, Bengio Y, Hinton G. Deep learning[J]. Nature, 2015, 521(7553): 436-444. doi: 10.1038/nature14539
[14]	Chen Y Y, Ming D P, Lü X W. Superpixel based land cover classification of VHR satellite image combining multi-scale CNN and scale parameter estimation[J]. Earth Science Informatics, 2019, 12(3): 341-363. doi: 10.1007/s12145-019-00383-2
[15]	Tong X Y, Xia G S, Lu Q, et al. Learning transferable deep models for land-use classification with high-resolution remote sensing images[J]. arXiv, 2018, 3: 05713, doi: 10.48550/arXiv.1807.05713.
[16]	Zhao W Z, Du S H. Learning multiscale and deep representations for classifying remotely sensed imagery[J]. Isprs Journal of Photogrammetry & Remote Sensing, 2016, 113: 155-165.
[17]	Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90. doi: 10.1145/3065386
[18]	Shin H C, Roth H R, Gao M, et al. Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning[J]. IEEE Transactions on Medical Imaging, 2016, 35(5): 1285-1298. doi: 10.1109/TMI.42
[19]	Bhowmick S, Nagarajaiah S, Veeraraghavan A. Vision and deep learning-based algorithms to detect and quantify cracks on concrete surfaces from UAV videos[J]. Sensors, 2020, 20(21): 6299, doi: 10.3390/s20216299.
[20]	Ronneberger O, Fischer P, Brox T. U-Net: Convolutional networks for biomedical image segmentation[C]// International Conference on Medical Image Computing and Computer-Assisted Intervention. Cham: Springer, 2015.
[21]	Li S J, Xiong L Y, Tang G A, et al. Deep learning-based approach for landform classification from integrated data sources of digital elevation model and imagery[J]. Geomorphology, 2020, 354: 107045, doi: 10.1016/j.geomorph.2020.107045.
[22]	陈晓征. 基于高精度DEM的黄土淤地坝信息提取及特征分析[D]. 南京: 南京师范大学, 2020.
	[ Chen Xiaozheng. Information extraction and feature analysis of loess check dam based on high-precision DEM[D]. Nanjing: Nanjing Normal University, 2020. ]
[23]	Yang X, Wang Y, Byrne R, et al. Concepts of artificial intelligence for computer-assisted drug discovery[J]. Chemical Reviews, 2019, 119(18): 10520-10594. doi: 10.1021/acs.chemrev.8b00728 pmid: 31294972
[24]	Li S J, Xiong L Y, Hu G H, et al. Extracting check dam areas from high-resolution imagery based on the integration of object-based image analysis and deep learning[J]. Land Degradation and Development, 2021, 32(7): 2303-2317. doi: 10.1002/ldr.v32.7
[25]	董斯齐, 黄翀. 粤港澳大湾区陆源氮污染来源结构与空间分布[J]. 环境科学, 2021, 42(11): 5384-5393. doi: 10.1021/es801743s
	[ Dong Siqi, Huang Chong. Source structure and spatial distribution of terrestrial nitrogen pollution in Guangdong Hong Kong Macao Greater Bay area[J]. Environmental Science, 2021, 42(11): 5384-5393. ] doi: 10.1021/es801743s
[26]	Stehman S V, Foody G M. Key issues in rigorous accuracy assessment of land cover products[J]. Remote Sensing of Environment, 2019, 231: 111199, doi: 10.1016/j.rse.2019.05.018.

基于深度学习融合OBIA的黄土高原小流域淤地坝系提取

Extraction of check dam system in small watershed of Loess Plateau based on deep learning and OBIA

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 26

相关文章 15

Metrics

本文评价

推荐阅读 0

提取方法	准确率	精准率	召回率	F1Score
深度学习融合OBIA方法	98.98	81.97	90.94	89.70
OBIA方法	96.67	96.52	52.20	67.76

[1]	任涛涛,李双双,段克勤,何锦屏. 黄土高原热浪型和缺水型骤旱时空变化特征及其影响因素[J]. 干旱区地理, 2023, 46(3): 360-370.
[2]	安彬,肖薇薇,刘宇峰,刘全玉. 1955—2021年黄土高原地区相对湿度时空演变规律[J]. 干旱区地理, 2023, 46(12): 1939-1950.
[3]	卓静,胡皓,何慧娟,王智,杨承睿. 陕北黄土高原生态脆弱性时空变异及驱动因素分析[J]. 干旱区地理, 2023, 46(11): 1768-1777.
[4]	沙国良,魏天兴,陈宇轩,傅彦超,任康. 黄土高原丘陵区典型植物群落土壤粒径分布特征[J]. 干旱区地理, 2022, 45(4): 1224-1234.
[5]	叶文丽,杨新军,吴孔森,王银. 黄土高原社会-生态系统恢复力时空变化特征与影响因素分析[J]. 干旱区地理, 2022, 45(3): 912-924.
[6]	吉珍霞,侯青青,裴婷婷,陈英,谢保鹏,吴华武. 黄土高原植被物候对季节性干旱的敏感性响应[J]. 干旱区地理, 2022, 45(2): 557-565.
[7]	孙健武,高军波,马志飞,喻超,张欣怡. 不同地理环境下“空间贫困陷阱”分异机制比较——基于大别山与黄土高原的实证[J]. 干旱区地理, 2022, 45(2): 650-659.
[8]	田达睿,唐皓,谭静斌. 陕北黄土高原丘陵沟壑区聚落适宜空间模式研究——以米脂县东沟为例[J]. 干旱区地理, 2022, 45(1): 263-276.
[9]	牛燕宁,綦琳,乔彦松. 基于碎屑锆石年代学对黄土高原物源及其时空差异的理解与展望[J]. 干旱区地理, 2021, 44(6): 1623-1634.
[10]	姜丽,魏天兴,李亦然,魏安琪. 地形因子对陕北黄土丘陵区防护林树种分布的影响[J]. 干旱区地理, 2021, 44(6): 1763-1771.
[11]	安彬,肖薇薇,张淑兰,朱妮,张建东. 1960—2017年黄土高原地表温度时空变化特征[J]. 干旱区地理, 2021, 44(3): 778-785.
[12]	王钧,李广,聂志刚,董莉霞,闫丽娟. 陇中黄土高原区旱地春小麦产量对干旱胁迫响应的模拟研究[J]. 干旱区地理, 2021, 44(2): 494-506.
[13]	姚雪玲,杨国靖,王帅,郭秀江. 黄土丘陵沟壑区不同深度土壤水分对降雨的响应及其稳定性[J]. 干旱区地理, 2021, 44(2): 507-513.
[14]	苗玉,高冠龙,李伟. 黄土高原苹果树叶片气孔导度的环境响应与模拟[J]. 干旱区地理, 2021, 44(2): 525-533.
[15]	李雨欣,薛东前,马蓓蓓,董朝阳. 黄土高原地区农村贫困空间演化及偏远特征[J]. 干旱区地理, 2021, 44(2): 534-543.