基于深度学习的滑坡监测与早期预警方法研究

详细信息    查看全文 | 推荐本文 |

英文篇名：Research on Landslide Monitoring and Early Warning Based on Depth Learning 作者：顾华奇 ; 陈皆红 ; 李婷 英文作者：GU Huaqi;CHEN Jiehong;LI Ting;Jiangxi Surveying and Mapping Geographic Information Engineering Technology Research Center;Jiangxi Basic Geographic Information Center;Jiangxi Normal University, Key Laboratory of Poyang Lake Wetland and Watershed Research,Ministry of Education;Jiangxi Provincial Key Laboratory of Poyang Lake Comprehensive Management and Resource Development,Jiangxi Normal University; 关键词：浅层滑坡 ; 大数据分析 ; 空间多源数据获取及融合 ; 卷积神经网络模型 ; 滑坡演变过程模拟 英文关键词：shallow landslide;;large data analysis;;spatial multi-source data acquisition and fusion;;convolutional neural network model;;Landslide Evolution Simulation 中文刊名：JSKX 英文刊名：Jiangxi Science 机构：江西省测绘地理信息工程技术研究中心;江西省基础地理信息中心;江西师范大学鄱阳湖湿地与流域研究教育部重点实验室地理与环境学院;江西师范大学江西省鄱阳湖综合治理与资源开发重点实验室; 出版日期：2019-04-15 出版单位：江西科学 年：2019 期：v.37;No.172 基金：江西省重点研发计划项目(编号:20171BBG70094);; 江西省生态安全协同创新课题(编号:JXS-EW-012);; 国家科技支撑计划课题项目(编号:2015BAH50FO1);; 国家自然科学基金项目(编号:41761076);; 江西省教育厅项目(编号:GJJ160319);; 鄱阳湖湿地与流域研究教育部重点实验室(江西师范大学)开放基金资助项目(编号:PK2017007) 语种：中文; 页：JSKX201902013 页数：6 CN：02 ISSN：36-1093/N 分类号：54-58+126

摘要

研究的浅层滑坡成灾机理及成灾模式分析、面向大数据分析的空间多源数据获取及融合、基于深度卷积神经网络模型的滑坡演变过程模拟、滑坡演化的定量评估模型和早期预警系统建设等4个有机部分,构成了一个从理论到方法再到应用的相对完整的技术主线。其中对浅层滑坡成灾机理及成灾模式分析为后三者的提供理论指导,面向大数据分析的空间多源数据获取及融合是基于深度卷积神经网络模型的滑坡演变过程模拟的数据基础,滑坡演化的定量评估模型和早期预警系统建设为技术路线的实现、验证与应用部分。
        The shallow landslide disaster mechanism and model analysis, spatial multi-source data acquisition and fusion for large data analysis, Landslide Evolution Process Simulation Based on deep convolution neural network model, quantitative assessment model of Landslide Evolution and early warning system construction, which are four organic parts of this paper, constitute an organic part. Relatively complete technical mainline from theory to method and then to application. Among them, the analysis of the mechanism and mode of shallow landslide disaster provides theoretical guidance for the latter three. The acquisition and fusion of spatial multi-source data for large data analysis is based on the data base of the simulation of the process of Landslide Evolution Based on the deep convolution neural network model. The quantitative evaluation model of Landslide Evolution and the construction of early warning system are the technologies. The realization, validation and application of surgical route.

引文

[1] 李德仁,廖明生,王艳.永久散射体雷达干涉测量技术[J].武汉大学学报(信息科学版),2004,29(8):664-668.
    [2] Ferretti A, Savio G, Barzaghi R, et al. Submillimeter accuracy of InSAR time series: experimental validation[J]. IEEE Geoscience and Remote Sensing, 2007, 45(5):1142-1153.
    [3] Zebker H A, Werner C L, Rosen P A, et al. Accuracy of topographic maps derived from ERS-1 interferometric radar[J]. IEEE Transactions on Geoscience and Remote Sensing, 1994, 32(4):823-836.
    [4] HanssenR F. Radar Interferometry: Data Interpretation and Error Analysis[M]. Dordrecht: Kluwer Academic Publishers, 2001.
    [5] Ferretti A, Prati C, Rocca F. Permanent scatterers in SAR interferometry[J]. IEEE Transactions on Geoscience and Remote Sensing, 2001, 39(1): 8-20.
    [6] Hooper A, Zebker H, Segall P, et a1. A new method for measuring deformation on volcanoes and other natural terrains using InSAR persistent scatterers[J]. Geophysical Research Letters, 2004, 31(23):147-159.
    [7] Hooper A, Segall P, Zebker H. Persistent scatterer InSAR for crustal deformation analysis, with application to volcan′ alcedo, galapagos[J]. Journal of Geophysical Research Atmospheres, 2007, 112(B7): 1-20.