Please wait a minute...
img

官方微信

遥感技术与应用  2021, Vol. 36 Issue (4): 865-872    DOI: 10.11873/j.issn.1004-0323.2021.4.0865
遥感应用     
基于多纹理特征融合的震后SAR图像倒塌建筑物信息提取
杜妍开(),龚丽霞(),李强,詹森,张景发
应急管理部自然灾害防治研究院,北京 100085
Earthquake Induced Building Damage Assessment on SAR Multi-texture Feature Fusion
Yankai Du(),Lixia Gong(),Qiang Li,Sen Zhan,Jingfa Zhang
National Institute of Natural Hazards,Beijing 100085,China
 全文: PDF(3939 KB)   HTML
摘要:

合成孔径雷达(SAR)凭借其全天候观测能力以及SAR图像中丰富的纹理信息,在震后建筑物倒塌评估中发挥了重要作用。针对SAR图像中倒塌建筑物纹理特征多样但利用率较低,且特征信息冗余的问题,提出一种基于主成分分析的SAR图像多纹理特征分类方法。该方法基于灰度直方图、灰度共生矩阵、局部二值模式、Gabor滤波器提取了26种纹理特征信息,构建主成分变量进行多维特征优选与降维融合,通过随机森林分类算法提取建筑物的倒塌信息。以2016年日本熊本地震为例验证了该方法的有效性,结果显示其提取精度高达79.85%,倒塌建筑物的识别效率有所提高,分类结果优于单种纹理特征提取方法及多种纹理特征组合提取法,可用于震后建筑物震害信息的快速提取。

关键词: 地震倒塌建筑物评估SAR多纹理特征主成分分析    
Abstract:

Synthetic Aperture Radar (SAR) plays an important role in building collapse assessment after earthquake with its all-weather observation capability and rich texture information in SAR images. In order to solve the problems of multi-texture features of collapsed buildings in SAR images, such as low utilization rate and redundant feature information, a multi-texture feature classification method based on Principal Component Analysis (PCA) is proposed. This method extracts 26 kinds of texture feature information based on gray-level histogram, gray level co-occurrence matrix, Local Binary Pattern (LBP) and Gabor filters, constructs principal component variable for multi-dimensional feature selection and dimension reduction fusion, and extracts collapse information of buildings through Random Forest classification algorithm. Taking the Kumamoto earthquake in Japan in 2016 as an example to verify the effectiveness of this method, the results show that the extraction accuracy is up to 79.85%, the identification efficiency of collapsed buildings is improved, and the classification results are superior to each texture feature extraction method and multi-texture feature combination extraction method, which can be used for the rapid extraction of earthquake damage information of buildings.

Key words: Earthquake    Building damage assessment    SAR    Multi-texture feature    Principal component analysis
收稿日期: 2020-05-19 出版日期: 2021-09-26
ZTFLH:  P237  
基金资助: 中国地震局地壳应力研究所中央级公益性科研院所基本科研业务专项资助项目(ZDJ2018-14)
通讯作者: 龚丽霞     E-mail: du_yankai@163.com;xiaolongzhu1900@hotmail.com
作者简介: 杜妍开(1995—),女,山西晋中人,硕士研究生,主要从事遥感震害评估研究。E?mail: du_yankai@163.com
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  
杜妍开
龚丽霞
李强
詹森
张景发

引用本文:

杜妍开,龚丽霞,李强,詹森,张景发. 基于多纹理特征融合的震后SAR图像倒塌建筑物信息提取[J]. 遥感技术与应用, 2021, 36(4): 865-872.

Yankai Du,Lixia Gong,Qiang Li,Sen Zhan,Jingfa Zhang. Earthquake Induced Building Damage Assessment on SAR Multi-texture Feature Fusion. Remote Sensing Technology and Application, 2021, 36(4): 865-872.

链接本文:

http://www.rsta.ac.cn/CN/10.11873/j.issn.1004-0323.2021.4.0865        http://www.rsta.ac.cn/CN/Y2021/V36/I4/865

图1  研究方法流程图
图2  基本LBP算子
图3  益城町地区遥感影像
图4  益城町地区震后建筑物目视解译结果
方法纹理特征提取精度/%
灰度直方图均值、方差、偏度、峰度、熵69.13
GLCM反差、相异度、同质性、角二阶矩、熵、GLCM均值、方差、相关性。74.06
LBP基本LBP、RILBP、ULBP69.75
Gabor滤波器G1、G2、G3、G4、G5、G6、G7、G874.94
所有特征组合75.93
表1  五种方法所用的纹理特征及提取精度
主分量序号特征值累积贡献率/%
11 881 537.2789.57
2175 685.8297.93
317 462.0698.76
415 293.7699.49
54 399.4099.70
62 548.5699.82
71 947.5999.92
81 118.2799.97
9595.78100.00
1058.27100.00
表2  前10个主成分的特征值和累积贡献率
图5  融合后的特征在不同维数下的分类精度
图6  建筑物震害信息提取结果图
震害类型验证数据总计UA/%
倒塌未倒塌
倒塌8 16614 29922 46536.35
未倒塌5 78255 17960 96190.52
总计13 94869 47883 426
PA/%58.5579.4275.93
表3  基于所有纹理特征提取精度分布
震害类型验证数据总计UA/%
倒塌未倒塌
倒塌10 93013 79424 72444.21
未倒塌3 01855 68458 70294.88
总计13 94869 47883 426
PA/%78.3680.1579.85
表4  基于主成分分析后的多纹理特征提取精度分布
特征维数训练时间/s提取精度/%
所有特征266675.93
主成分分析44179.85
表5  基于两种方法分类效率精度对比
1 Xue Tengfei. Extraction of seismic buildings based on SAR multi-feature change detection [D].Harbin:Institute of Engineering Mechanics,China Earthquake Administration,2017.薛腾飞.基于SAR多特征变化检测的震害建筑物提取研究.哈尔滨:中国地震局工程力学研究所,2017.
2 Natsuaki R, Nagai H, Tomii N, et al. Sensitivity and limitation in damage detection for individual buildings using InSAR Coherence-a case study in 2016 Kumamoto earthquakes[J]. Remote Sensing,2018(10):245. DOI:10.3390/rs10020245.
doi: 10.3390/rs10020245
3 Zhan Sen, Zhang Jingfa, Wang Jianfei, et al. Earthquake induced building damage extraction based on multi-temporal and dual-polarized Sentinel-1A data[J]. Journal of Natural Disasters,2019,28(5):220-228.
3 詹森,张景发,王建飞,等.基于多时相双极化Sentinel-1A数据的震害建筑物提取[J].自然灾害学报,2019,28(5):220-228.
4 Guida R, Lodice A, Riceio D. An application of the deterministic feature extraction approach to COSMO-SKyMed data[C]∥ Proceedings of the 2010 8th European Conference on Synthetic Aperture Radar (EUSAR). Aachen,Germany:VDE,2010:1-4.
5 Brett P T B. Urban damage detection in high resolution amplitude images[D]. Surrey:Surrey Space Centre, University of Surrey,2013.
6 Liu Jinyu, Zhang Jingfa, Liu Guolin. An analysis of earthquake damage information based on imaging mechanism of the high resolution SAR image[J]. Remote Sensing for Land and Resources,2013,25(3):61-65.
6 刘金玉,张景发,刘国林. 基于高分辨率SAR图像成像机理的震害信息分析[J].国土资源遥感,2013,25(3):61-65.
7 Cui Liping, Wang Xiaoqing, Dou Aixia, et al. Building damage analysis based on high resolution Synthetic Aperture Radar imaging geometry[J].Earthquake Science,2016,38(2):272-282.
7 崔丽萍,王晓青,窦爱霞,等. 基于高分辨率合成孔径雷达影像建筑物成像几何结构的震害特征分析[J].地震学报,2016,38(2):272-282.
8 Li Qiang, Zhang Jingfa, Gong Lixia, et al. Extraction of earthquake-collapsed buildings based on correlation change detection of multi-texture features in SAR images[J]. Journal of Remote Sensing,2018,22(Sup.1):128-138.
8 李强,张景发,龚丽霞,等. SAR图像纹理特征相关变化检测的震害建筑物提取[J].遥感学报,2018,22():128-138.
9 Guo Huadong, Wang Xinyuan, Li Xinwu, et al. Yushu earthquake synergic analysis study using multi-model SAR datasets[J]. Chinese Science Bulletin,2010,55(31):1195-1199.
9 郭华东,王心源,李新武,等. 多模式SAR玉树地震协同分析[J].科学通报,2010,55(13):1195-1199.
10 Sato M, Chen S W. Detection of damaged area by polarimetric SAR[C]∥ Proeedings of the 2013 Asia-Pacific Conference on Synthetic Aperture Radar(APSAR).Tsukuba,Japan:IEEE,2013:451-454.
11 Karimzadeh S, Matsuoka M. Building damage characterization for the 2016 amatrice earthquake using ascending descending COSMO-SkyMed data and topographic position index[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018(8):1-15.DOI: 10.1109/JSTARS.2018.2825399.
doi: 10.1109/JSTARS.2018.2825399
12 Zhai Wei, Shen Huanfeng, Huang Chunlin. Collapsed buildings extraction from the PolSAR image based on the analysis of texture features[J]. Remote Sensing Technology and Application,2016,31(5):975-982.
12 翟玮,沈焕锋,黄春林.结合PolSAR影像纹理特征分析提取倒塌建筑物[J].遥感技术与应用,2016,31(5):975-982.
13 Yaqi J, Josaphat S S, Ming C, et al. Earthquake/Tsunami damage assessment for urban areas using post-event PolSAR data[J]. Remote Sensing, 2018, 10(7):1088. DOI: 10.3390/rs10071088.
doi: 10.3390/rs10071088
14 Romaniello V, Piscini A, Bignami C, et al. Earthquake damage mapping by using remotely sensed data: the Haiti case study[J]. Journal of Applied Remote Sensing, 2017,11(1):016042. DOI: 10.1117/1.JRS.11.016042.
doi: 10.1117/1.JRS.11.016042
15 Shi L, Sun W, Yang J, et al. Building collapse assessment by the use of post-earthquake Chinese VHR airborne SAR[J]. IEEE Geoscience&Remote Sensing Letters,2015,12(10):2021-2025. DOI: 10.1109/LGRS.2015.2443018.
doi: 10.1109/LGRS.2015.2443018
16 Wu F, Gong L, Wang C, et al. Signature analysis of building damage with TerraSAR-X new staring spotlight mode data[J]. IEEE Geoscience and Remote Sensing Letters, 2016(99):1-5. DOI: 10.1109/LGRS.2016.2604841.
doi: 10.1109/LGRS.2016.2604841
17 Dell'Acqua F, Polli D A. Postrevent only VHR Radar satellite data for automated damage assessment[J]. Photogrammetric Engineering Remote Sensing,2011,77(10):1037-1043.
18 Gong L X, Wang C, Wu F, et al. Earthquake induced building damage detection with post-event submeter VHR terraSAR-X staring spotlight imagery[J]. Remote Sensing,2016,8(12):887. DOI: 10.3390/rs8110887.
doi: 10.3390/rs8110887
19 Chen Qihao, Nie Yuliang, Li Linlin, et al. Buildings damage assessment using texture features of polarization decomposition components[J]. Journal of Remote Sensing, 2017, 21(6): 955-965.
19 陈启浩,聂宇靓,李林林,等.极化分解后多纹理特征的建筑物损毁评估[J].遥感学报, 2017, 21(6):955-965.
20 Ge P, Gokon H, Meguro K. Building damage assessment using intensity SAR data with different incidence angles and longtime interval[J]. Journal of Disaster Research,2019,14(3):456-465. DOI: 10.20965/jdr.2019.p0456.
doi: 10.20965/jdr.2019.p0456
21 Sun W, Shi L, Yang J, et al. Building collapse assessment in urban areas using texture information from post-event SAR data[J]. IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing,2016,9(8):3792-3808. DOI: 10.1109/JSTARS.2016.2580610.
doi: 10.1109/JSTARS.2016.2580610
22 Ojala T, Pietikainen M, Maenpaa T. Multiresolution gray-scale and rotation invariant texture classification with local binary patterns[C]∥ IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(7):971-987.
23 Sandler R, Lindenbaum M. Gabor filter analysis for texture segmentation[C]∥ Computer Vision and Pattern Recognition Workshop,CVPRW '06.
23 Conference on. IEEE,2006.
24 Zhang Bin, Gao Xin. A Multi-feature joint SAR texture image classification method based on Gabor filter bank and LBP[J]. Science Technology and Engineering,2010,10(17):4196-4201.
24 张斌,高鑫.一种基于Gabor滤波器组和LBP的多特征联合SAR纹理图像分类方法[J].科学技术与工程,2010,10(17):4196-4201.
25 Deng Lei, Li Jing, Nie Juan. A method fusing SAR with multi-spectral image and reducing speckle noise[J].Journal of Natural Disasters,2008,17(6):87-90.
25 邓磊,李京,聂娟,孙洪泉.抑制斑点噪声的SAR与多光谱图像融合方法[J].自然灾害学报,2008,17(6):87-90.
26 Chen Miaojin, Wang Xiaoqin, Wu Siying. Importance analysis of soil erosion influencing factors based on random forest [J].Journal of Natural Disasters,2019,28(4):209-219.
26 陈妙金,汪小钦,吴思颖.基于随机森林算法的水土流失影响因子重要性分析[J].自然灾害学报,2019,28(4):209-219.
27 Zhan Sen, Zhang Jingfa, Gong Lixia, et al. Seismic damage identification of single buildings in high resolution Synthetic Aperture radar image based on texture features[J]. Science Technology and Engineering,2019,19(31):47-54.
27 詹森,张景发,龚丽霞,等.基于纹理特征的高分辨率合成孔径雷达影像单体建筑物震害信息识别[J].科学技术与工程,2019,19(31):47-54.
28 Li Q, Gong L, Zhang J. A correlation change detection method integrating PCA and multi-texture features of SAR image for building damage detection[J]. European Journal of Remote Sensing,2019,52(1):435-447. DOI:10.1080/22797254. 2019.1630322.
doi: 10.1080/22797254. 2019.1630322
[1] 陈子涵,王峰,许宁,尤红建. 基于改进NSST-PCNN的光学与SAR图像融合去云方法[J]. 遥感技术与应用, 2021, 36(4): 810-819.
[2] 安炳琪,罗海滨,丁海勇,张志山,王伟,史潇,柯福阳,王明明. 基于SBAS-InSAR技术的西宁地表形变监测[J]. 遥感技术与应用, 2021, 36(4): 838-846.
[3] 李昕娟,林家元,胡桂胜,赵伟. 西南山地典型流域地震前后泥石流物源遥感精细识别[J]. 遥感技术与应用, 2021, 36(3): 638-648.
[4] 王晨丞,王永前,王利花. 基于SAR纹理信息的农作物识别研究——以农安县为例[J]. 遥感技术与应用, 2021, 36(2): 372-380.
[5] 金亚秋. 空间微波遥感研究与应用—丛书述评[J]. 遥感技术与应用, 2021, 36(1): 1-10.
[6] 周金霖,陈锟山,古琼昇,曾江源,许镇. 基于多次散射目标模型的雷达成像研究[J]. 遥感技术与应用, 2021, 36(1): 132-140.
[7] 任慧敏,宋冬梅,王斌,甄宗晋,刘斌,张婷. 基于新极化特征参数的SAR海洋溢油检测[J]. 遥感技术与应用, 2020, 35(4): 934-942.
[8] 肖修来,翟玮,郭晓,裴万胜,邓津. 极化SAR建筑物震害信息识别研究方法综述[J]. 遥感技术与应用, 2020, 35(3): 509-516.
[9] 王树果, 马春锋, 赵泽斌, 魏龙. 基于Sentinel-1及Landsat 8数据的黑河中游农田土壤水分估算[J]. 遥感技术与应用, 2020, 35(1): 13-22.
[10] 杨燕,李震,黄磊,田帮森. 高分辨率SAR影像提取冰川面积与冰面河[J]. 遥感技术与应用, 2019, 34(6): 1155-1161.
[11] 李春江,沈国状,张继超. 基于灰色系统理论的植被物理参数与极化分解参数的关联分析[J]. 遥感技术与应用, 2019, 34(4): 839-846.
[12] 苏华,张明慧,李静,陈修治,汪小钦. 基于光学与SAR因子的森林生物量多元回归估算[J]. 遥感技术与应用, 2019, 34(4): 847-856.
[13] 徐梦竹, 徐佳, 邓鸿儒, 袁春琦. 基于全极化SAR影像的海岛地物分类[J]. 遥感技术与应用, 2019, 34(3): 647-654.
[14] 李春江, 沈国状, 张继超. 基于灰色系统理论的植被物理参数与极化分解参数的关联分析—以鄱阳湖湿地为例[J]. 遥感技术与应用, 2019, 34(2): 284-292.
[15] 宋小霞, 王静, 储小青. 基于多普勒频移的SAR海表流场反演[J]. 遥感技术与应用, 2019, 34(2): 293-302.