基于GEE和U-net模型的同震滑坡识别方法

doi:10.12082/dqxxkx.2022.210704

地球信息科学学报 ›› 2022, Vol. 24 ›› Issue (7): 1275-1285.doi: 10.12082/dqxxkx.2022.210704

• 地球信息科学理论与方法 • 上一篇下一篇

基于GEE和U-net模型的同震滑坡识别方法

刘佳¹^,³(), 伍宇明¹^,^*(), 高星¹, 司文涛²

1.中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室,北京 100101
2.山东师范大学公共管理学院,济南 250100
3.中国科学院大学,北京 100049

收稿日期:2021-11-03 修回日期:2022-01-17 出版日期:2022-07-25 发布日期:2022-09-25
通讯作者: * 伍宇明（1989—）,北京人,男,博士,副研究员,主要从事地质灾害研究,E-mail: wuym@lreis.ac.cn
作者简介:刘佳（1993—）,男,河北沧州人,博士研究生,主要研究方向为滑坡识别与动力学模拟,E-mail: liuj.19b@igsnrr.ac.cn
基金资助:
中国科学院战略性先导科技专项（A类）子课题(XDA23090503);国家重点研发计划项目(YS2018YFGH000001);国家重点研发计划项目(2018YFC1505503)

Image Recognition of Co-seismic Landslide based on GEE and U-net Neural Network

LIU Jia¹^,³(), WU Yuming¹^,^*(), GAO Xing¹, SI Wentao²

1. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
2. School of Public Administration, Shandong Normal University, Jinan 250100, China
3. University of Chinese Academy of Sciences, Beijing 100049, China

Received:2021-11-03 Revised:2022-01-17 Online:2022-07-25 Published:2022-09-25
Contact: WU Yuming
Supported by:
The Strategic Priority Research Program of the Chinese Academy of Sciences(XDA23090503);National Key Research and Development Plan of China(YS2018YFGH000001);National Key Research and Development Plan of China(2018YFC1505503)

摘要/Abstract

摘要：

地震滑坡解译是震后重建的重要基础工作,主要通过室内人工遥感解译和室外野外调查确定。地震滑坡相比其他地物来说更为复杂,很难通过简单指数识别。室内遥感解译通过滑坡后壁、侧壁和堆积等纹理特征进行识别,大面积同震滑坡解译工作往往耗费大量人力和物力,且耗时长,难以满足灾害应急需求。本研究利用U-net神经网络模型,结合Google Earth Engine（GEE）云平台和人工智能学习平台Tensorflow,以地震局解译的汶川滑坡作为样本数据,以震后30 m分辨率的Landsat影像、高程、坡度以及NDVI数据作为模型输入参数,自动识别并获取了汶川地震后的同震滑坡数据,同时比较了不同参数组合情况下U-net神经网络模型的分割识别精度。研究表明：① U-net模型可以用于以Landsat影像为基础数据的同震滑坡快速自动识别;② 随着高程、坡度以及NDVI等输入参数增加,模型分割精度在逐渐提高,但假阳性结果也会出现增多,震后滑坡影像+高程+坡度+NDVI的输入参数组合精度最高;③ 在细节上,模型在多参数组合的情况下,大型滑坡能够很好被识别,一些较小型的滑坡受制于影像分辨率的影响,分割精度较差。为了更好识别小型滑坡,后续研究可能需提高影像的分辨率。此外,GEE云平台大大提高了训练样本获取的效率,为科研人员快速进行基于神经网络与遥感数据的地物识别研究提供了条件。

关键词: 地震滑坡, U-net模型, GEE, 图像分割, Landsat影像, 均方误差, 召回率, 平均交并比

Abstract:

Landslide identification is a fundamental work in geological survey. Characterizing the distribution of earthquake-induced landslides can help define where areas are prone to landslides and where areas are safe, which is important for subsequent post-earthquake reconstruction. At present, the co-seismic landslide is usually identified by remote sensing interpretation and field survey based on color changes and textual characteristics such as crown cracks, accumulation, and flank. However, the identification and segmentation of large-area earthquake-induced landslides often takes a lot of time and resources, making it difficult to provide timely relevant data to rescuers in disasters. Recently, the combination of Google Earth Engine (GEE), Colab, and Cloud storage provides new technique support for post-earthquake landslide interpretation. Remote sensing images can be quickly extracted in GEE, learned in the Colab platform, and stored in Cloud storage. Based on these technologies, a U-net model built on Tensorflow was integrated in the GEE to identify co-seismic landslides of Wenchuan. We tested three combinations of different remote sensing data to identify landslides, namely Landsat 5 (LT05), LT05+slope+DEM, and LT05+slope+DEM+NDVI. The results showed that all these three combinations had an accuracy of 0.71~0.87, precision of 0.66~0.75, recall of 0.29~0.57, and mIoU of 0.50~0.64. Due to the complexity of landslides in remote sensing images, the U-net model still resulted in misclassification and cannot replace human interpretation completely. However, this model can help us in interpreting earthquake-induced landslides to some extent. From the detailed images, there is a high possibility for areas that have been marked as landslides, but there may also be a large number of landslides for areas that are not marked as landslides. The accuracy and precision increased about 1% in all three data combinations with the increasing number of parameters, but the number of false positives also increased with the increasing number of parameters. The NDVI, a derivative image from LT05, enhanced learning efficiency and accuracy, which indicated that the prior knowledge was very important in a neural network model. By comparison, the model can segment large-scale landslides well in the case of using the combination of LT05+NDVI+DEM+slope. However, due to the limited image resolution, the segmentation accuracy of small landslides was still not good. We need higher spatial resolution images to improve segmentation accuracy. This study demonstrates the computation efficiency of the GEE+Colab+Cloud storage platform, which provides a technique support for researchers to conduct object recognition quickly using remote sensing data based on neural networks.

Key words: co-seismic landslide, U-net model, GEE, image segmentation, Landsat, MSE, recall, mIoU

刘佳, 伍宇明, 高星, 司文涛. 基于GEE和U-net模型的同震滑坡识别方法[J]. 地球信息科学学报, 2022, 24(7): 1275-1285.DOI:10.12082/dqxxkx.2022.210704

LIU Jia, WU Yuming, GAO Xing, SI Wentao. Image Recognition of Co-seismic Landslide based on GEE and U-net Neural Network[J]. Journal of Geo-information Science, 2022, 24(7): 1275-1285.DOI:10.12082/dqxxkx.2022.210704

图/表 10

图1

表1

图2

图3

图4

表2

表3

表4

表5

表6

参考文献 26

[1]	Chen C W, Tung Y S, Liou J J, et al. Assessing landslide characteristics in a changing climate in northern Taiwan[J]. CATENA, 2019, 175:263-277. DOI: 10.1016/j.catena.2018.12.023 doi: 10.1016/j.catena.2018.12.023
[2]	黄润秋. 中国西部地区典型岩质滑坡机理研究[J]. 地球科学进展, 2004, 19(3):443-450.
	[ Huang R Q. Mechanism of large scale landslides in Western China[J]. Advance in Earth Sciences, 2004, 19(3):443-450. ] DOI: 10.3321/j.issn:1001-8166.2004.03.016 doi: 10.3321/j.issn:1001-8166.2004.03.016
[3]	吕杰堂, 王治华, 周成虎. 西藏易贡大滑坡成因探讨[J]. 地球科学, 2003, 28(1):107-110.
	[ Lü J T, Wang Z H, Zhou C H. Discussion on the occurrence of yigong landslide in Tibet[J]. Earth Science, 2003, 28(1):107-110. ] DOI: 10.3321/j.issn:1000-2383.2003.01.018 doi: 10.3321/j.issn:1000-2383.2003.01.018
[4]	许强, 郑光, 李为乐, 等. 2018年10月和11月金沙江白格两次滑坡-堰塞堵江事件分析研究[J]. 工程地质学报, 2018, 26(6):1534-1551.
	[ Xu Q, Zheng G, Li W L. et al. Study on successive landslide damming events of Jinsha River in baige village on octorber 11 and November 3, 2018[J]. Journal of Engineering Geology, 2018, 26(6):1534-1551. ]DOI: 10.13544/j.cnki.jeg.2018-406 doi: 10.13544/j.cnki.jeg.2018-406
[5]	冯文凯, 张国强, 白慧林, 等. 金沙江“10·11”白格特大型滑坡形成机制及发展趋势初步分析[J]. 工程地质学报, 2019, 27(2):415-425.
	[ Feng W K, Zhang G Q, Bai H L. et al. A preliminary analysis of the formation mechanism and development tendency of the huge baige landslide in Jinsha River on October 11, 2018[J]. Journal of Engineering Geology, 2019, 27(2):415-425. ]DOI: 10.13544/j.cnki.jeg.2018-392 doi: 10.13544/j.cnki.jeg.2018-392
[6]	Yin Y P, Wang F W, Sun P. Landslide hazards triggered by the 2008 Wenchuan earthquake, Sichuan, China[J]. Landslides, 2009, 6(2):139-152. DOI: 10.1007/s10346-009-0148-5. doi: 10.1007/s10346-009-0148-5
[7]	Dai F C, Xu C, Yao X, et al. Spatial distribution of landslides triggered by the 2008 Ms 8.0 Wenchuan earthquake, China[J]. Journal of Asian Earth Sciences, 2011, 40(4):883-895. DOI: 10.1016/j.jseaes.2010.04.010 doi: 10.1016/j.jseaes.2010.04.010
[8]	王治华, 吕杰堂. 从卫星图像上认识西藏易贡滑坡[J]. 遥感学报, 2001, 5(4):312-316.
	[ Wang Z H, Lü J T. Understand yigong landslide in Tibet based on the satellite image[J]. Journal of Remote Sensing, 2001, 5(4):312-316. ]
[9]	姚鑫, 邓建辉, 刘星洪, 等. 青藏高原泛三江并流区活动性滑坡InSAR初步识别与发育规律分析[J]. 工程科学与技术, 2020, 52(5):16-37.
	[ Yao X, Deng J H, Liu X H, et al. Primary recognition of active landslides and development rule analysis for pan three-river-parallel territory of Tibet Plateau[J]. Advanced Engineering Sciences, 2020, 52(5):16-37. ] DOI: 10.15961/j.jsuese.202000529 doi: 10.15961/j.jsuese.202000529
[10]	Guo X Y, Guo Q, Feng Z K. Detecting the vegetation change related to the creep of 2018 baige landslide in Jinsha River, SE Tibet using SPOT data[J]. Frontiers in Earth Science, 2021, 9:706998. DOI: 10.3389/feart.2021.706998 doi: 10.3389/feart.2021.706998
[11]	Yang W T, Wang Y Q, Sun S, et al. Using Sentinel-2 time series to detect slope movement before the Jinsha River landslide[J]. Landslides, 2019, 16(7):1313-1324. DOI: 10.1007/s10346-019-01178-8 doi: 10.1007/s10346-019-01178-8
[12]	Ni, Yang, Li, et al. Decreasing trend of geohazards induced by the 2008 Wenchuan earthquake inferred from time series NDVI data[J]. Remote Sensing, 2019, 11(19):2192. DOI: 10.3390/rs11192192 doi: 10.3390/rs11192192
[13]	Ji L, Zhang L, Wylie B. Analysis of dynamic thresholds for the normalized difference water index[J]. Photogrammetric Engineering & Remote Sensing, 2009, 75(11):1307-1317. DOI: 10.14358/pers.75.11.1307 doi: 10.14358/pers.75.11.1307
[14]	李勋, 杨环, 殷宗敏, 等. 基于DEM和遥感影像的区域黄土滑坡体识别方法研究[J]. 地理与地理信息科学, 2017, 33(4):86-92,129.
	[ Yang H, Yin Z M, et al. Regional loess landslide recognition method research based on DEM and remote sensing image[J]. Geography and Geo-Information Science, 2017, 33(4):86-92,129. ] DOI: 10.3969/j.issn.1672-0504.2017.04.015 doi: 10.3969/j.issn.1672-0504.2017.04.015
[15]	彭令, 徐素宁, 梅军军, 等. 地震滑坡高分辨率遥感影像识别[J]. 遥感学报, 2017, 21(4):509-518.
	[ Peng L, Xu S N, Mei J J, et al. Earthquake-induced landslide recognition using high-resolution remote sensing images[J]. Journal of Remote Sensing, 2017, 21(4):509-518. ] DOI: 10.11834/jrs.20176176 doi: 10.11834/jrs.20176176
[16]	Ghorbanzadeh O, Blaschke T, Gholamnia K, et al. Evaluation of different machine learning methods and deep-learning convolutional neural networks for landslide detection[J]. Remote Sensing, 2019, 11(2):196. DOI: 10.3390/rs11020196 doi: 10.3390/rs11020196
[17]	Qin S W, Guo X, Sun J B, et al. Landslide detection from open satellite imagery using distant domain transfer learning[J]. Remote Sensing, 2021, 13(17):3383. DOI: 10.3390/rs13173383 doi: 10.3390/rs13173383
[18]	Meena S R, Ghorbanzadeh O, Westen C J, et al. Rapid mapping of landslides in the Western Ghats (India) triggered by 2018 extreme monsoon rainfall using a deep learning approach[J]. Landslides,2021, 18(5):1937-1950. DOI: 10.1007/s10346-020-01602-4 doi: 10.1007/s10346-020-01602-4
[19]	Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation[C]// 2015 IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 2015:3431-3440. DOI: 10.1109/CVPR.2015.7298965 doi: 10.1109/CVPR.2015.7298965
[20]	Ronneberger O, Fischer P, Brox T. U-net: Convolutional networks for biomedical image segmentation[M]// Lecture Notes in Computer Science. Cham: Springer International Publishing, 2015:234-241. DOI: 10.1007/978-3-319-24574-4_28 doi: 10.1007/978-3-319-24574-4_28
[21]	Yu B, Chen F, Xu C. Landslide detection based on contour-based deep learning framework in case of national scale of Nepal in 2015[J]. Computers & Geosciences, 2020, 135:104388. DOI: 10.1016/j.cageo.2019.104388 doi: 10.1016/j.cageo.2019.104388
[22]	Yi Y N, Zhang W C. A new deep-learning-based approach for earthquake-triggered landslide detection from single-temporal RapidEye satellite imagery[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020, 13:6166-6176. DOI: 10.1109/JSTA RS.2020.3028855 doi: 10.1109/JSTA RS.2020.3028855
[23]	Qi S, Xu Q, Lan H, et al. Spatial distribution analysis of landslides triggered by 2008.5. 12 Wenchuan Earthquake, China[J]. Engineering Geology, 2020, 116(1-2):95-108. DOI: 10.1016/j.enggeo.2010.07.011 doi: 10.1016/j.enggeo.2010.07.011
[24]	Xu C, Xu X W, Yao X, et al. Three (nearly) complete inventories of landslides triggered by the May 12, 2008 Wenchuan Mw 7.9 earthquake of China and their spatial distribution statistical analysis[J]. Landslides, 2014, 11(3):441-461. DOI: 10.1007/s10346-013-0404-6 doi: 10.1007/s10346-013-0404-6
[25]	田乃满, 兰恒星, 伍宇明, 等. 人工神经网络和决策树模型在滑坡易发性分析中的性能对比[J]. 地球信息科学学报, 2020, 22(12):2304-2316. doi: 10.12082/dqxxkx.2020.190766
	[ Tian N M, Lan H X, Wu Y M, et al. Performance comparison of BP artificial neural network and CART decision tree model in landslide susceptibility prediction[J]. Journal of Geo-information Science, 2020, 22(12):2304-2316.] DOI: 10.12082/dqxxkx.2020.190766 doi: 10.12082/dqxxkx.2020.190766
[26]	Ji S P, Yu D W, Shen C Y, et al. Landslide detection from an open satellite imagery and digital elevation model dataset using attention boosted convolutional neural networks[J]. Landslides, 2020, 17(6):1337-1352. DOI: 10.1007/s103 46-020-01353-2 doi: 10.1007/s103 46-020-01353-2

数据类型	数据说明	获取日期	数据来源
滑坡数据	汶川地震滑坡数据集（Xu）	2008年5月13日至2014年	USGS网站
震后影像	Landsat-5 1-7波段数据	2008年5月13日至2010年12月31日	GEE平台
高程数据	SRTM 90 m分辨率数据	2000年	GEE平台
坡度数据	基于高程数据计算得到	2000年	GEE平台
NDVI数据	根据震后影像数据计算得到	2008年5月13日至2010年12月31日	GEE平台

		A区实际	B区实际				C区实际
		P	N	P	N	P	N
方案1预测	P	4731	1774	9836	4211	5232	2572
（LT05）	N	8929	50 102	14 603	36 886	7115	50 617
方案2预测	P	4022	1337	10 669	4830	6579	2960
（LT05+DEM+SLOPE）	N	9638	50 539	13 770	36 267	5768	50 229
方案3预测	P	4868	1796	12 141	6060	7066	3241
（LT05+DEM+SLOPE+NDVI）	N	8792	50 080	12 298	35 037	5281	49 948

	A区	B区	C区
LT05	0.73	0.70	0.67
LT05+DEM+SLOPE	0.75	0.69	0.69
LT05+NDVI+DEM+SLOPE	0.73	0.67	0.69

	A区	B区	C区
LT05	0.84	0.71	0.85
LT05+DEM+SLOPE	0.83	0.72	0.87
LT05+NDVI+DEM+SLOPE	0.84	0.72	0.87

	A区	B区	C区
LT05	0.35	0.40	0.42
LT05+DEM+SLOPE	0.29	0.44	0.53
LT05+NDVI+DEM+SLOPE	0.36	0.50	0.57

基于GEE和U-net模型的同震滑坡识别方法

Image Recognition of Co-seismic Landslide based on GEE and U-net Neural Network

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 26

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	巫磊, 吴文挺. GEE平台下结合滤波算法和植被物候特征的互花米草遥感提取最优时间窗口确定[J]. 地球信息科学学报, 2023, 25(3): 606-624.
[2]	尹雄, 陈帮乾, 古晓威, 云挺, 吴志祥, 陈岳, 赖虹燕, 寇卫利. 基于GEE平台LandTrendr算法的海南岛森林扰动快速监测方法及分析[J]. 地球信息科学学报, 2023, 25(10): 2093-2106.
[3]	张歆越, 高晓路, 柴琪, 宋敦江. 中国地级城市建成区蓝绿空间时空格局及其气候影响因素分析[J]. 地球信息科学学报, 2023, 25(1): 190-207.
[4]	张旭, 柳林, 周翰林, 岳瀚, 孙秋远. 基于贝叶斯逻辑回归模型研究百度街景图像微观建成环境因素对街面犯罪的影响[J]. 地球信息科学学报, 2022, 24(8): 1488-1501.
[5]	陈振, 陈芸芝, 吴婷, 李佳优. 面向高分遥感影像道路提取的轻量级双注意力和特征补偿残差网络模型[J]. 地球信息科学学报, 2022, 24(5): 949-961.
[6]	詹琪琪, 赵伟, 杨梦娇, 付浩, 李昕娟, 熊东红. 雅鲁藏布江中部流域土地沙化遥感识别[J]. 地球信息科学学报, 2022, 24(2): 391-404.
[7]	乔丹玉, 郑进辉, 鲁晗, 邓磊. 面向不同环境背景的Landsat影像水体提取方法适用性研究[J]. 地球信息科学学报, 2021, 23(4): 710-722.
[8]	冯叶涵, 陈亮, 贺晓冬. 基于百度街景的SVF计算及其在城市热岛研究中的应用[J]. 地球信息科学学报, 2021, 23(11): 1998-2012.
[9]	何红术, 黄晓霞, 李红旮, 倪凌佳, 王新歌, 陈崇, 柳泽. 基于改进U-Net网络的高分遥感影像水体提取[J]. 地球信息科学学报, 2020, 22(10): 2010-2022.
[10]	何昭欣, 张淼, 吴炳方, 邢强. Google Earth Engine支持下的江苏省夏收作物遥感提取[J]. 地球信息科学学报, 2019, 21(5): 752-766.
[11]	汤冬梅, 樊辉, 张瑶. Landsat时序变化检测综述[J]. 地球信息科学学报, 2017, 19(8): 1069-1079.
[12]	林齐根, 刘燕仪, 刘连友, 王瑛. 支持向量机与Newmark模型结合的地震滑坡易发性评估研究[J]. 地球信息科学学报, 2017, 19(12): 1623-1633.
[13]	苏腾飞, 张圣微, 李洪玉. 基于优化合并的高分辨率遥感影像分割算法[J]. 地球信息科学学报, 2016, 18(7): 931-940.
[14]	王志华, 孟樊, 杨晓梅, 杨丰硕, 方豫. 高空间分辨率遥感影像分割尺度参数自动选择研究[J]. 地球信息科学学报, 2016, 18(5): 639-648.
[15]	殷瑞娟, 施润和, 李镜尧. 一种高分辨率遥感影像的最优分割尺度自动选取方法[J]. 地球信息科学学报, 2013, 15(6): 902-910.