基于Deeplab V3+的2019年长宁M6.0地震建筑物震害信息提取研究

doi:10.12196/j.issn.1000-3274.2022.03.009

摘要/Abstract

摘要： 快速获取地震所造成的建筑物震害信息是灾后救援及灾害损失评估的关键, 以语义分割网络为代表的深度学习技术的发展, 为震后建筑物震害信息的获取提供了一种新的手段。本文将Deeplab V3+神经网络模型应用于建筑物震害信息的提取, 以2019年长宁M6.0地震震后双河镇0.06 m分辨率的无人机影像作为建筑物震害信息提取试验数据源, 将试验区建筑物划分为无明显破坏、破坏两类, 其余作为背景信息, 人工进行建筑物震害信息的分类与标注, 构建样本数据集并进行网络模型的训练。为了提高模型针对不同场景下地物类别分类的鲁棒性与泛化能力, 进行了样本增强处理, 开展了样本增强对模型精度的影响实验与训练样本数量对模型精度的影响实验, 并以九寨沟漳扎镇震后无人机0.4 m分辨率影像进行模型的可迁移性实验。实验结果表明, 本次研究所建立的神经网络模型在建筑物震害信息的提取中可以达到较高的提取精度; 样本增强后较样本增强前模型各类别提取精度都有不同程度的提高, 特别是破坏建筑物召回率提高约0.15, 精确率提高约0.20; 该神经网络模型在未做任何修改的情况下, 具有一定的可迁移性。

关键词: 深度学习, Deeplab V3+, 建筑物, 震害提取, 样本增强

Abstract: Rapid acquisition of building damage information caused by earthquakes is the key to post-disaster rescue and disaster loss assessment. The development of deep learning technology represented by semantic segmentation network provides a new means for the acquisition of building damage information after earthquakes. In this paper, Deeplab V3 + neural network model is applied to the extraction of earthquake damage information of buildings. The UAV image with 0.06 meters resolution in Shuanghe Town after Chang Ning earthquake is used as the data source for the extraction of earthquake damage information of buildings. The buildings in the test area are divided into two categories: no obvious damage and damage, and the rest are used as background information. The seismic damage information of buildings is classified and annotated manually, and the sample data set is constructed and the network model is trained. In order to improve the robustness and generalization ability of the model for the classification of ground objects in different scenarios, the sample enhancement process was carried out, and the experiment of the effect of sample enhancement on the model accuracy and the effect of the number of training samples on the model accuracy were carried out. After the earthquake, the 0.4-meter-resolution UAV image was used to conduct the model transferability experiment. The experimental results show that the neural network model established in this study can achieve higher extraction accuracy in the extraction of building earthquake damage information; the extraction accuracy of each category of the model after sample enhancement is improved to varying degrees, especially it is that the recall rate of the destroyed building is increased by about 0.15, and the precision rate is increased by about 0.20; the neural network model has a certain transferability without any modification.

Key words: Deep learning, Deeplab V3+, Building, Earthquake damage extraction, Sample enhancement

中图分类号:

TP79

王鑫, 窦爱霞, 郭红梅, 袁小祥. 基于Deeplab V3+的2019年长宁M6.0地震建筑物震害信息提取研究[J]. 地震, 2022, 42(3): 124-140.

WANG Xin, DOU Ai-xia, GUO Hong-mei, YUAN Xiao-xiang. Research on Extraction of Building Damage Information in the 2019 Changning M6.0 Earthquake Based on Deeplab V3+[J]. EARTHQUAKE, 2022, 42(3): 124-140.

参考文献

[1] 李继业. 震灾评估仿真系统的研究与实现[D]. 上海: 复旦大学, 2010.
LI Ji-ye. Research and realization of earthquake disaster assessment simulation system[D]. Shanghai: Fudan University, 2010 (in Chinese).
[2] 王晓青, 窦爱霞, 王龙, 等. 2013年四川芦山7.0级地震烈度遥感评估[J]. 地球物理学报, 2015, 58(1): 163-171.
WANG Xiao-qing, DOU Ai-xia, WANG Long, et al. RS-based assessment of seismic intensity of the 2013 Lushan, Sichuan, China M_S7.0 earthquake[J]. Chinese Journal of Geophysics, 2015, 58(1): 163-171 (in Chinese).
[3] 于洋, 施国武, 刘斌, 等. 基于全卷积神经网络的无人机影像建筑物提取[J]. 水利水电技术, 2020, 51(7): 31-38.
YU Yang, SHI Guo-wu, LIU Bin, et al. Fully convolutional network-based building extraction of image from unmanned aerial vehicle[J]. Water Resources and Hydropower Engineering, 2020, 51(7): 31-38 (in Chinese).
[4] 张春森, 葛英伟, 蒋萧. 基于稀疏约束SegNet的高分辨率遥感影像建筑物提取[J]. 西安科技大学学报, 2020, 40(3): 441-448.
ZHANG Chun-sen, GE Ying-wei, JIANG Xiao. High-resolution remote sensing image building extraction based on sparsely constrained SegNet[J]. Journal of Xi'an University of Science and Technology, 2020, 40(3): 441-448 (in Chinese).
[5] 张浩然, 赵江洪, 张晓光. 利用U-Net网络的高分遥感影像建筑提取方法[J]. 遥感信息, 2020, 35(3): 143-150.
ZHANG Hao-ran, ZHAO Jiang-hong, ZHANG Xiao-guang. High-resolution image building extraction using U-net Neural Network[J]. Remote Sensing Information, 2020, 35(3): 143-150 (in Chinese).
[6] 焦利伟, 张敏, 麻连伟, 等. 基于Pytorch框架搭建U-Net网络模型的遥感影像建筑物提取研究[J]. 河南城建学院学报, 2020, 29(4): 52-57.
JIAO Li-wei, ZHANG Min, MA Lian-wei, et al. Building extraction from remote sensing images using U-Net network model based on Pytorch framework[J]. Journal of Henan University of Urban Construction, 2020, 29(4): 52-57 (in Chinese).
[7] 杨乐, 王慧, 李烁, 等. 结合DeepLabv3架构的多源数据建筑物提取方法[J]. 测绘与空间地理信息, 2020, 43(6): 62-66.
YANG Le, WANG Hui, LI Shuo, et al. Multi-source data building extraction method combined with DeepLabv3 architecture[J]. Geomatics & Spatial Information Technology, 2020, 43(6): 62-66 (in Chinese).
[8] Chen L C, Zhu Y K, Papandreou G, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[J]. Computer Vision and Pattern Recognition, 2018.1802.02611.
[9] 陈丹丹. 基于Deeplabv3+的高分遥感影像道路损毁信息提取方法研究[D]. 北京: 中国地震局地震预测研究所, 2020.
CHEN Dan-dan. Research on the extraction method of road damage information from high-resolution remote sensing images based on Deeplabv3+[D]. Beijing: Institute of Earthquake Forecasting, China Earthquake Administration, 2020 (in Chinese).
[10] Arbib M A. The handbook of brain theory and neural networks[M]∥Lecun Y, Bengio Y. Convolutional networks for images, speech and time series. Cambridge: MIT Press, 1995: 255-258.
[11] Chen L C, Papandreou G, Kokkinos I, et al. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848.
[12] He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition[C]∥2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR), June 27-30, 2016, Las Vegas, NV, USA. New York: IEEE, 2016: 770-778.
[13] Yu F, Koltun V. Multi-scale context aggregation by dilated convolutions[J/OL]. (2016-04-30) [2020-10-15].
[14] Cheng X Y, Zhao L Z, Hu Q, et al. Real-time semantic segmentation based on dilated convolution smoothing and lightweight up-sampling[J]. Laser & Optoelectronics Progress, 2020, 57(2): 021017.
[15] 施国武, 邢宽平, 张俊贤, 等. 基于深度学习的无人机影像建筑物自动提取[J]. 地矿测绘, 2020, 36(1): 28-31.
SHI Guo-wu, XING Kuan-ping, ZHANG Jun-xian, et al. Buildings extraction from UAV images based on deep learning[J]. Surveying and Mapping of Geology and Mineral Resources, 2020, 36(1): 28-31 (in Chinese).
[16] Cubuk E D, Zoph B, Mane D, et al. Auto augment: Learning augmentation policies from data[J]. Computer Vision and Pattern Recognition, 2018, 10(3): 108-124.
[17] Chawla N V, Bowyer K W, Hall L O, et al. SMOTE: synthetic minority over-sampling technique[J]. Journal of Artificial Intelligence Research, 2002, 16(1): 321-357.
[18] Inoue H. Data Augmentation by Pairing Samples for Images Classification[J]. IEEE Geoscience & Remote Sensing Letters, 12(11): 2321-2325.
[19] Zhang H Y, Cisse M, Dauphin Y N, et al. Mixup: Beyond empirical risk minimization[J]. International Journal of Remote Sensing, 36(12): 3144-3169.
[20] 王波, 郭迅, 宣越. 基于新视角的震害分析以四川长宁6.0级地震为例[J]. 震灾防御技术, 2020, 15(3): 496-509.
WANG Bo, GUO Xun, XUN Yue. Analysis of earthquake damage from a new perspective: A case study of Changning M6.0 earthquake in Sichuan Province[J]. Technology for Earthquake Disaster Prevention, 2020, 15(3): 496-509 (in Chinese).
[21] 中国地震台网中心. 中国地震台网历史查询[EB/OL]. (2019-06-24)[2019-09-04]. http:∥www.ceic.ac.cn/history.
China Earthquake Network Center. History query of China Earthquake Networks[EB/OL]. (2019-06-24)[2019-09-04]. http:∥www.ceic.ac.cn/history (in Chinese).
[22] 应急管理部. 应急管理部持续指导四川长宁地震抗震救灾工作[EB/OL]. (2019-06-18)[2019-09-04]. https:∥www.mem.gov.cn/xw/bndt/201906/t20190618_310848.shtml.
Ministry of Emergency Management. The Ministry of Emergency Management continues to guide the earthquake relief work in Changning, Sichuan[EB/OL]. (2019-06-18)[2019-09-04]. https:∥www.mem.gov.cn/xw/bndt/201906/t20190618_310848.shtml (in Chinese).
[23] 潘毅, 陈建, 包韵雷, 等. 长宁6.0级地震村镇建筑震害调查与分析[J]. 建筑结构学报, 2020, 41(S1): 297-306.
PAN Yi, CHEN Jian, BAO Yun-lei, et al. Investigation and analysis of earthquake damage to villages and towns in Changning M6.0 earthquake[J]. Journal of Building Structures, 2020, 41(S1): 297-306 (in Chinese).
[24] 陈梦, 王晓青. 全卷积神经网络在建筑物震害遥感提取中的应用研究[J]. 震灾防御技术, 2019, 14(4): 810-820.
CHEN Meng, WANG Xiao-qing. The study on extraction of seismic damage of buildings from remote sensing image based on fully convolutional neural network[J]. Technology for Earthquake Disaster Prevention, 2019, 14(4): 810-820 (in Chinese).