基于LiDAR和航空影像的地震灾害倒塌建筑物信息提取
|
摘要
地震灾害损失评估是震后展开救灾工作的重要环节。快速、准确地获取震后损毁建筑物信息能够为灾区减灾、救灾工作提供有效的支持。高分辨率航空遥感是灾害监测的重要技术手段,但其信息自动提取的精度受到一定的限制。近年来新出现的LiDAR技术能够提供地面目标的高程信息,可应用于复杂环境下倒塌建筑物信息的提取。研究中采用航空遥感数据和LiDAR数据,基于面向对象的图像分析(Object-Based Image Analysis,OBIA)与SVM技术相结合的方法对2010年1月12日海地地震中倒塌建筑物信息进行了提取,提取总体精度达到86.1%。
Damage estimation caused by an earthquake is a major task in the post-disaster mitigation process.To enhance the relief and rescue operation in the affected area,it is required to receive rapid and accurate knowledge about the conditions of damaged area.Remote sensing techniques were proved to be useful in the last decades in detecting,identifying and monitoring the impact and effect of natural disasters.Recently emerging LiDAR data provide the height of the ground objects,which can be used to extract the collapsed building in a complex urban environment.Using the aerophotographs and the normalized digital surface model(nDSM) extracted from LiDAR data,the authors developed a method based on OBIA and SVM for extracting the earthquake-caused collapsed building.The test study in Port-au-Prince,Haiti's capital,after January 12,2010 earthquake shows that the method can extract collapsed buildings with high accuracy of 86.1%.
Damage estimation caused by an earthquake is a major task in the post-disaster mitigation process.To enhance the relief and rescue operation in the affected area,it is required to receive rapid and accurate knowledge about the conditions of damaged area.Remote sensing techniques were proved to be useful in the last decades in detecting,identifying and monitoring the impact and effect of natural disasters.Recently emerging LiDAR data provide the height of the ground objects,which can be used to extract the collapsed building in a complex urban environment.Using the aerophotographs and the normalized digital surface model(nDSM) extracted from LiDAR data,the authors developed a method based on OBIA and SVM for extracting the earthquake-caused collapsed building.The test study in Port-au-Prince,Haiti's capital,after January 12,2010 earthquake shows that the method can extract collapsed buildings with high accuracy of 86.1%.
引文
[1]郭华东,鹿琳琳,马建文,等.一种改进的地震灾害倒塌房屋遥感信息自动识别方法[J].科学通报,2009,54(17):2581-2585.
[2]Kaya S,Curran P J,Llewellyn G.Post-earthquake BuildingCollapse:A Comparison of Government Statistics and EstimatesDerived from SPOT HRVIR Data[J].Int J Remote Sens,2005,26(3):2731-2740.
[3]Sakamoto M,Takasago Y,Uto K,et al.Automatic Detection ofDamaged Area of Iran Earthquake by High-resolution SatelliteImagery[C]∥Proceedings of IGARSS’04.Alaska,2004:1418-1421.
[4]Turker M,San B T.Detection of Collapsed Buildings Caused by the1999 Izmit,Turkey Earthquake Through Digital Analysis of Post-event Aerial Photographs[J].Int J Remote Sens,2004,25(21):4701-4714.
[5]Turker M,Cetinkaya B.Automatic Detection of Earthquake-damaged Buildings Using DEMs Created from Pre-and Post-earthquake Stereo Aerial Photographs[J].Int J Remote Sens,20052,6(4):823-832.
[6]Gamba P,Dell’Acqua F,Trianni G.Rapid Damage Detection inthe Bam Area Using Multitemporal SAR and Exploiting AncillaryData[J].IEEE Transactions on Geoscience&Remote Sensing,20074,5(6):1582-1589.
[7]Alexander B,Christian H,Goepfert J,et al.Aspects of GeneratingPrecise Digital Terrain Models in the Wadden Sea from Lidar-water Classification and Structure Line Extraction[J].ISPRSJournal of Photogrammetry&Remote Sensing,20086,3(5):510-528.
[8]Axelsson P.DEM Generation from Laser Scanner Data UsingAdaptive TIN Models[J].International Archives of Photogramme-try and Remote Sensing,20003,3(1):110-117.
[9]Baatz M,Schpe A.Multiresolution Segmentation—an OptimizationApproach for High Quality Multi-scale Image Segmentation[C]∥Strobl J,Blaschke T,Griesebner G.Angewandte GeographischeInformations-Verarbeitung XII.Karlsruhe:Wichmann Verlag,2000:12-23.
[10]Haralick R M,Shanmugan K,Dinstein I.Textural Features forImage Classification[J].IEEE Transactions on Systems,Man,andCybernetics1,973,3(6):610-621.
[11]于海洋,甘甫平,武法东,等.VHSR图像基于分割对象分类器性能评价[J].国土资源遥感2,008(2):30-34.
[12]Foody G M,Mathur A.Toward Intelligent Training of SupervisedImage Classifications:Directing Training Data Acquisition for SVMClassification[J].Remote Sensing of Environment,2004,93(1/2):107-117.
[13]Vapnik V.The Nature of Statistical Learning Theory[M].NewYork:Springer-Verlag,1995.
[2]Kaya S,Curran P J,Llewellyn G.Post-earthquake BuildingCollapse:A Comparison of Government Statistics and EstimatesDerived from SPOT HRVIR Data[J].Int J Remote Sens,2005,26(3):2731-2740.
[3]Sakamoto M,Takasago Y,Uto K,et al.Automatic Detection ofDamaged Area of Iran Earthquake by High-resolution SatelliteImagery[C]∥Proceedings of IGARSS’04.Alaska,2004:1418-1421.
[4]Turker M,San B T.Detection of Collapsed Buildings Caused by the1999 Izmit,Turkey Earthquake Through Digital Analysis of Post-event Aerial Photographs[J].Int J Remote Sens,2004,25(21):4701-4714.
[5]Turker M,Cetinkaya B.Automatic Detection of Earthquake-damaged Buildings Using DEMs Created from Pre-and Post-earthquake Stereo Aerial Photographs[J].Int J Remote Sens,20052,6(4):823-832.
[6]Gamba P,Dell’Acqua F,Trianni G.Rapid Damage Detection inthe Bam Area Using Multitemporal SAR and Exploiting AncillaryData[J].IEEE Transactions on Geoscience&Remote Sensing,20074,5(6):1582-1589.
[7]Alexander B,Christian H,Goepfert J,et al.Aspects of GeneratingPrecise Digital Terrain Models in the Wadden Sea from Lidar-water Classification and Structure Line Extraction[J].ISPRSJournal of Photogrammetry&Remote Sensing,20086,3(5):510-528.
[8]Axelsson P.DEM Generation from Laser Scanner Data UsingAdaptive TIN Models[J].International Archives of Photogramme-try and Remote Sensing,20003,3(1):110-117.
[9]Baatz M,Schpe A.Multiresolution Segmentation—an OptimizationApproach for High Quality Multi-scale Image Segmentation[C]∥Strobl J,Blaschke T,Griesebner G.Angewandte GeographischeInformations-Verarbeitung XII.Karlsruhe:Wichmann Verlag,2000:12-23.
[10]Haralick R M,Shanmugan K,Dinstein I.Textural Features forImage Classification[J].IEEE Transactions on Systems,Man,andCybernetics1,973,3(6):610-621.
[11]于海洋,甘甫平,武法东,等.VHSR图像基于分割对象分类器性能评价[J].国土资源遥感2,008(2):30-34.
[12]Foody G M,Mathur A.Toward Intelligent Training of SupervisedImage Classifications:Directing Training Data Acquisition for SVMClassification[J].Remote Sensing of Environment,2004,93(1/2):107-117.
[13]Vapnik V.The Nature of Statistical Learning Theory[M].NewYork:Springer-Verlag,1995.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心