基于航空影像和LiDAR数据的海地地震滑坡识别研究
|
摘要
快速、准确地判别滑坡灾害范围对灾情评估与救灾指挥至关重要。将高分辨率航空影像和机载LiDAR数据相结合,能充分利用滑坡体的光谱特征、几何形态与空间结构特征,提高滑坡识别的准确度。该文以海地地震诱发的滑坡识别为例,介绍了面向对象的滑坡识别思路及方法;针对海地滑坡灾害特点,选择了相应的灾害特征,并经反复试验设定了特征参数阈值;针对面向对象分类过程中特征组合与选择这一瓶颈问题,以河流沉积物和滑坡判别为例,通过特征组合试验与对比分析,发现基于NDVI与坡度特征组合的分类结果明显优于单一特征及其它组合模式的分类结果。与最大似然分类法相比,该方法的分类结果更为准确、精度更高。
An earthquake measuring 7.0 struck southern Haiti on 12 January,2010,causing tremendous damage in the Port-au-Prince area and in southern Haiti,killing or injuring thousands of people and knocking down a large number of buildings,as well as quantities of landslides.Fast and reliable detection of earthquake-induced landslides is crucial for disaster reduction and emergency management to limit life losses and properties.Integration of aerial images and LiDAR data is capable of utilizing sufficiently the spectral characteristics,surface texture and spatial structural features to improve the accuracy of landslides identification.Taking the case of Haiti earthquake induced landslides,an object-based method was developed in this paper to identify earthquake-induced landslides from aerial images and LiDAR data,which contains noise and small details generally.The data sets,data pre-processing,multi-scale image segmentation,the selection of feature and its threshold,rules construction,the combination of features and its trials were presented.With the accuracy and precision analysis,it had proved that the proposed method is better than the maximum likelihood method for earthquake-induced landslides detection.Experimental results indicate that the classification of fluvial deposit and earthquake-induced landslides with the feature combination of NDVI and slope value is better than that with NDVI only or other combinations.The researches demonstrate the feasibility and effectiveness of the proposed approach on practical applications.
An earthquake measuring 7.0 struck southern Haiti on 12 January,2010,causing tremendous damage in the Port-au-Prince area and in southern Haiti,killing or injuring thousands of people and knocking down a large number of buildings,as well as quantities of landslides.Fast and reliable detection of earthquake-induced landslides is crucial for disaster reduction and emergency management to limit life losses and properties.Integration of aerial images and LiDAR data is capable of utilizing sufficiently the spectral characteristics,surface texture and spatial structural features to improve the accuracy of landslides identification.Taking the case of Haiti earthquake induced landslides,an object-based method was developed in this paper to identify earthquake-induced landslides from aerial images and LiDAR data,which contains noise and small details generally.The data sets,data pre-processing,multi-scale image segmentation,the selection of feature and its threshold,rules construction,the combination of features and its trials were presented.With the accuracy and precision analysis,it had proved that the proposed method is better than the maximum likelihood method for earthquake-induced landslides detection.Experimental results indicate that the classification of fluvial deposit and earthquake-induced landslides with the feature combination of NDVI and slope value is better than that with NDVI only or other combinations.The researches demonstrate the feasibility and effectiveness of the proposed approach on practical applications.
引文
[1]王治华.中国滑坡遥感及新进展[J].国土资源遥感,2007,74(4):7-10.
[2]SASSA K,FFKUOKA H,WANG F,et al.Progress in Land-slide Science[M].Berlin:Springer-Verlag,2007.
[3]KEEPER D K.Landslides caused by earthquakes[J].GeologicalSociety of America Bulletin,1984,95(4):406-421.
[4]OFDA/CRED.EM-DAT International Disaster Database-www.em-dat.net[R].UniversitéCatholique de Louvain,Brussels,Bel-gium,2006.
[5]CASSON B,DELACOURT C,BARATOUX D,et al.Seventeenyears of the"La Clapiere"landslide evolution analysed from or-tho-rectified aerial photographs[J].Engineering Geology,2003,68:123-139.
[6]NICHOL J,WONG M S.Satellite remote sensing for detailed land-slide inventories using change detection and image fusion[J].Inter-national Journal of Remote Sensing,2005,26:1913-1926.
[7]张学儒,刘林山,张镱锂,等.基于ENVI ZOOM面向对象的高海拔灌丛植被提取——以定日县为例[J].地理与地理信息科学,2010,26(4):104-108.
[8]SHEN Y L,SHENG Y H,ZHANG K,et al.Feature extractionfrom vehicle-borne laser scanning data[A].The InternationalConference on Earth Observation Data Processing and Analysis[C].Proc.SPIE,2008,7285:72850R.
[9]GLENN N F,STREUTKER D R,CHADWICK D J,et al.A-nalysis of LiDAR-derived topographic information for character-izing and differentiating landslide morphology and activity[J].Geomorphology,2006,73:131-148.
[10]MARTHA T R,KERLE N,JETTEN V,et al.Characterisingspectral,spatial and morphometric properties of landslides forsemi-automatic detection using object-oriented methods[J].Geomorphology,2010,116(1-2):24-36.
[11]MYNENI R B,HALL F G,SELLERS P J,et al.The interpre-tation of spectral vegetation indexes[J].IEEE Transactions onGeoscience and Remote Sensing,1995,33:481-486.
[12]eCognition,USER GUIDE.Definiens Imaging GmbH,Munich[EB/OL].www.ecognition.com,2010-08-30.
[13]SITHOLE G,VOSSELMAN G.Experimental comparison offilter algorithms for bare earth extraction from airborne laserscanning point clouds[J].ISPRS Journal of Photogrammetryand Remote Sensing,2004,59(1-2):85-101.
[14]ELMQVIST M.Ground estimation of lasar radar data usingactive shape models[A].The OEEPE Workshop on AirborneLaserscanning and Interferometric SAR for Detailed DigitalElevation Models[C].Stockholm,Sweden,2001.
[15]PFEIFER N,KOSTLI A,KRAUS K.Interpolation and filte-ring of laser scanner data-implementation and first results[A].International Archives of Photogrammetry and Remote Sens-ing[C].Columbus,1998,Vol XXXII,part 3/1,153-159.
[16]付卓,胡吉平,谭衢霖,等.遥感应用分析中影像分割方法[J].遥感技术与应用,2006,21(5):456-462.
[17]苏理宏,李小文,黄裕霞.遥感尺度问题研究进展[J].地球科学进展,2001,16(4):544-548.
[18]田新光.面向对象高分辨率遥感影像信息提取[D].中国测绘科学研究院,2007.
[19]郑毅,武法东,刘艳芳.一种面向对象分类的特征分析方法[J].地理与地理信息科学,2010,26(2):19-22.
[2]SASSA K,FFKUOKA H,WANG F,et al.Progress in Land-slide Science[M].Berlin:Springer-Verlag,2007.
[3]KEEPER D K.Landslides caused by earthquakes[J].GeologicalSociety of America Bulletin,1984,95(4):406-421.
[4]OFDA/CRED.EM-DAT International Disaster Database-www.em-dat.net[R].UniversitéCatholique de Louvain,Brussels,Bel-gium,2006.
[5]CASSON B,DELACOURT C,BARATOUX D,et al.Seventeenyears of the"La Clapiere"landslide evolution analysed from or-tho-rectified aerial photographs[J].Engineering Geology,2003,68:123-139.
[6]NICHOL J,WONG M S.Satellite remote sensing for detailed land-slide inventories using change detection and image fusion[J].Inter-national Journal of Remote Sensing,2005,26:1913-1926.
[7]张学儒,刘林山,张镱锂,等.基于ENVI ZOOM面向对象的高海拔灌丛植被提取——以定日县为例[J].地理与地理信息科学,2010,26(4):104-108.
[8]SHEN Y L,SHENG Y H,ZHANG K,et al.Feature extractionfrom vehicle-borne laser scanning data[A].The InternationalConference on Earth Observation Data Processing and Analysis[C].Proc.SPIE,2008,7285:72850R.
[9]GLENN N F,STREUTKER D R,CHADWICK D J,et al.A-nalysis of LiDAR-derived topographic information for character-izing and differentiating landslide morphology and activity[J].Geomorphology,2006,73:131-148.
[10]MARTHA T R,KERLE N,JETTEN V,et al.Characterisingspectral,spatial and morphometric properties of landslides forsemi-automatic detection using object-oriented methods[J].Geomorphology,2010,116(1-2):24-36.
[11]MYNENI R B,HALL F G,SELLERS P J,et al.The interpre-tation of spectral vegetation indexes[J].IEEE Transactions onGeoscience and Remote Sensing,1995,33:481-486.
[12]eCognition,USER GUIDE.Definiens Imaging GmbH,Munich[EB/OL].www.ecognition.com,2010-08-30.
[13]SITHOLE G,VOSSELMAN G.Experimental comparison offilter algorithms for bare earth extraction from airborne laserscanning point clouds[J].ISPRS Journal of Photogrammetryand Remote Sensing,2004,59(1-2):85-101.
[14]ELMQVIST M.Ground estimation of lasar radar data usingactive shape models[A].The OEEPE Workshop on AirborneLaserscanning and Interferometric SAR for Detailed DigitalElevation Models[C].Stockholm,Sweden,2001.
[15]PFEIFER N,KOSTLI A,KRAUS K.Interpolation and filte-ring of laser scanner data-implementation and first results[A].International Archives of Photogrammetry and Remote Sens-ing[C].Columbus,1998,Vol XXXII,part 3/1,153-159.
[16]付卓,胡吉平,谭衢霖,等.遥感应用分析中影像分割方法[J].遥感技术与应用,2006,21(5):456-462.
[17]苏理宏,李小文,黄裕霞.遥感尺度问题研究进展[J].地球科学进展,2001,16(4):544-548.
[18]田新光.面向对象高分辨率遥感影像信息提取[D].中国测绘科学研究院,2007.
[19]郑毅,武法东,刘艳芳.一种面向对象分类的特征分析方法[J].地理与地理信息科学,2010,26(2):19-22.
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