基于多光谱影像辅助的微波遥感水体提取方法研究
|
摘要
微波遥感数据具有穿透力强的特点,可在灾后复杂天气条件下快速获取灾区地表信息,如水体面积。但是在山区等复杂地形下,由于成像机理造成的阴影,会影响到水体提取的精度。为了快速有效地去除阴影,本文利用数据融合的方法,开展了基于光学影像辅助的微波遥感水体提取方法研究,并结合COSMO Skymed SAR数据和福卫-2号多光谱数据开展了实验分析。通过对HSV、Brovey、主成分、Gram-Schmidt四种融合方法效果的比较,发现无论从目视判读还是在定量指标上,Gram-Schmidt方法的效果都好于其他方法。将融合后图像进行监督分类,可以有效地区分出大部分雷达阴影和水体,从而快速有效地辅助在雷达图像上获取水体信息。
Microwave remote sensing has stronger penetration,which make it possible to obtain land surface information under the complex weather condition.But radar shadow caused by mountains will affect the accuracy of water extraction.In order to quickly and effectively distinguish the shadow,this paper uses the method of image fusion to extract water based on microwave images and multi-spectral images by using the COSMO Skymed SAR and Formosat-2 multi-spectral satellite data.Comparing four fusion methods(HSV,PCA,Gram-Schmidt,Brovey) whether by visual interpretation or in terms of quantitative indicators,the result shows that Gram-Schmidt method is a better fusion method,which can distinguish water greatly from other surface features.This paper presents that most of radar shadow and water can be separated efficiently by supervised classification on the fused image,which can be an important auxiliary information to extract the water area from the SAR data.
Microwave remote sensing has stronger penetration,which make it possible to obtain land surface information under the complex weather condition.But radar shadow caused by mountains will affect the accuracy of water extraction.In order to quickly and effectively distinguish the shadow,this paper uses the method of image fusion to extract water based on microwave images and multi-spectral images by using the COSMO Skymed SAR and Formosat-2 multi-spectral satellite data.Comparing four fusion methods(HSV,PCA,Gram-Schmidt,Brovey) whether by visual interpretation or in terms of quantitative indicators,the result shows that Gram-Schmidt method is a better fusion method,which can distinguish water greatly from other surface features.This paper presents that most of radar shadow and water can be separated efficiently by supervised classification on the fused image,which can be an important auxiliary information to extract the water area from the SAR data.
引文
[1]王治华,徐起德,杨日红,等.中印边界附近帕里河上的滑坡灾害遥感调查[J].科技导报,2007,25(6):27-31.
[2]Quincey D J,Richardson S D,Luckman A,et al.Early recognition of glacial lake hazards in the Himalaya usingremote sensing datasets[J].Global and Planetary Change.2007,56(1-2):137-152.
[3]Xu M,Cao C,Zhang H,et al.Change detection of an earthquake-induced barrier lake based on remote sensingimage classification[J].International Journal of Remote Sensing.2010,31(13):3521-3534.
[4]杨存建,魏一鸣.基于DEM的SAR图像洪水水体的提取[J].自然灾害学报,2002,11(3):121-125.
[5]史旭,周扬.汶川地震堰塞湖的雷达遥感识别研究[J].水电能源科学,2009,27(5):65-68.
[6]杨存建,周成虎.利用RADARSAT SWA SAR和LANDSAT TM的互补信息确定洪水水体范围[J].自然灾害学报,2001,10(2):79-83.
[7]陆渊,林怡,张绍明.灰度阈值方法下的SAR图像水体检测[J].山东建筑大学学报,2010(2):125-128.
[8]李秉柏,杨沈斌,申双和.ENVISAT ASAR数据与CBERS-02数据的融合与应用[J].地球信息科学,2008,10(2):263-268.
[9]陈东,李飚,沈振康.SAR与可见光图像融合算法的研究[J].系统工程与电子技术,2000,22(9):5-7,43.
[10]COSMO-Skymed SAR Products Handbook[R].Italian Space Agency COSMO-Skymed Mission,2009.
[11]Carper W,Lillesand T,Kiefer R.The use of intensity-hue-saturation transformations for merging SPOT panchro matic and multispectral image data[J].Photogrammetric Engineering and remote sensing,1990,56:459-467.
[12]Chavez P,Sides S C,Anderson J A.Comparison of three different methods to merge multiresolution and multi spectral data-Landsat TM and SPOT panchromatic[J].Photogrammetric Engineering and remote sensing,1991,57(3):295-303.
[13]北京大学数学系.高等代数(第二版)[M].北京:高等数学出版社,1998.
[14]崔岩梅,倪国强,钟堰利,王毅,钮永胜.利用统计特性进行图像融合效果分析及评价[J].北京理工大学学报,2000,20(1):102-106.
[2]Quincey D J,Richardson S D,Luckman A,et al.Early recognition of glacial lake hazards in the Himalaya usingremote sensing datasets[J].Global and Planetary Change.2007,56(1-2):137-152.
[3]Xu M,Cao C,Zhang H,et al.Change detection of an earthquake-induced barrier lake based on remote sensingimage classification[J].International Journal of Remote Sensing.2010,31(13):3521-3534.
[4]杨存建,魏一鸣.基于DEM的SAR图像洪水水体的提取[J].自然灾害学报,2002,11(3):121-125.
[5]史旭,周扬.汶川地震堰塞湖的雷达遥感识别研究[J].水电能源科学,2009,27(5):65-68.
[6]杨存建,周成虎.利用RADARSAT SWA SAR和LANDSAT TM的互补信息确定洪水水体范围[J].自然灾害学报,2001,10(2):79-83.
[7]陆渊,林怡,张绍明.灰度阈值方法下的SAR图像水体检测[J].山东建筑大学学报,2010(2):125-128.
[8]李秉柏,杨沈斌,申双和.ENVISAT ASAR数据与CBERS-02数据的融合与应用[J].地球信息科学,2008,10(2):263-268.
[9]陈东,李飚,沈振康.SAR与可见光图像融合算法的研究[J].系统工程与电子技术,2000,22(9):5-7,43.
[10]COSMO-Skymed SAR Products Handbook[R].Italian Space Agency COSMO-Skymed Mission,2009.
[11]Carper W,Lillesand T,Kiefer R.The use of intensity-hue-saturation transformations for merging SPOT panchro matic and multispectral image data[J].Photogrammetric Engineering and remote sensing,1990,56:459-467.
[12]Chavez P,Sides S C,Anderson J A.Comparison of three different methods to merge multiresolution and multi spectral data-Landsat TM and SPOT panchromatic[J].Photogrammetric Engineering and remote sensing,1991,57(3):295-303.
[13]北京大学数学系.高等代数(第二版)[M].北京:高等数学出版社,1998.
[14]崔岩梅,倪国强,钟堰利,王毅,钮永胜.利用统计特性进行图像融合效果分析及评价[J].北京理工大学学报,2000,20(1):102-106.
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