南方典型植被遥感丰度信息提取
|
摘要
针对遥感影像在南方丘陵地区典型植被丰度信息提取中存在的大量混合像元问题,为进一步提高线性解混精度,通过计算像元EVI值,构建了Landsat 8时间序列影像南方典型植被(端元)和混合像元的EVI时间序列曲线,分析了不同生育期内各种地物类型的植被指数变化曲线,发现不同地物在植被指数时间序列中具有各自独立的波动规律。选取多个端元及其EVI时间序列曲线,利用光谱匹配方法对匹配EVI时间序列曲线和多个端元进行了匹配,达到利用不同端元组合进行光谱解混的目的。试验结果表明,与传统方法相比,阔叶林解混精度有明显提高,针叶林解混精度也有所提高。该研究成果可以为南方丘陵地区植被环境的研究提供有力支撑。
Aiming at the problem of a large number of mixed pixels in the extraction of typical vegetation abundance in the southern hilly region of remote sensing images,in order to further improve the precision of linear unmixing,the EVI time series curves of typical vegetation( endmember) and mixed pixels in the south of Landsat 8 time series images are constructed by calculating the EVI value of pixels. The vegetation index change curves of various feature types in different growth periods are analyzed,and it is found that different features have their own independent fluctuation rules in the vegetation index time series. By selecting multiple endmembers and their EVI time series curves,the spectral matching method is used to match the EVI time series curve and multiple endmembers,and the purpose of spectral unmixing using different endmember combinations is achieved. The test results show that compared with the traditional method,the precision of broad-leaf forest unmixing is obviously improved,and the precision of coniferous forest unmixing is also improved. The research results can provide strong support for the study of vegetation environment in the southern hilly region.
Aiming at the problem of a large number of mixed pixels in the extraction of typical vegetation abundance in the southern hilly region of remote sensing images,in order to further improve the precision of linear unmixing,the EVI time series curves of typical vegetation( endmember) and mixed pixels in the south of Landsat 8 time series images are constructed by calculating the EVI value of pixels. The vegetation index change curves of various feature types in different growth periods are analyzed,and it is found that different features have their own independent fluctuation rules in the vegetation index time series. By selecting multiple endmembers and their EVI time series curves,the spectral matching method is used to match the EVI time series curve and multiple endmembers,and the purpose of spectral unmixing using different endmember combinations is achieved. The test results show that compared with the traditional method,the precision of broad-leaf forest unmixing is obviously improved,and the precision of coniferous forest unmixing is also improved. The research results can provide strong support for the study of vegetation environment in the southern hilly region.
引文
[1]李苗苗,吴炳方,颜长珍,等.密云水库上游植被覆盖度的遥感估算[J].资源科学,2004,26(4):153-159.
[2] DYMOND J R,STEPHENS P R,NEWSOME P F,et al.Percentage vegetation cover of a degrading rangeland from SPOT[J]. International Journal of Remote Sensing,1992,13(11):1999-2007.
[3]王圆圆,李京.遥感影像土地利用/覆盖分类方法研究综述[J].遥感信息,2004,19(1):53-59.
[4]罗红霞,龚健雅.线性和非线性光谱混合模型模拟土壤、植被混合光谱的效果分析[J].测绘通报,2005(5):6-10.
[5]王天星,陈松林,马娅.基于改进线性光谱分离模型的植被覆盖度反演[J].地球信息科学,2008,10(1):114-120.
[6] RIDD M K. Exploring a V-I-S(vegetation-impervious surfacesoil)model for urban ecosystem analysis through remote sensing:comparative anatomy for cities?[J]. International Journal of Remote Sensing,1995,16(12):2165-2185.
[7]李巍岳,胡志斌,闫巧玲.基于ETM~+影像的森林资源信息提取———以黄土高原丘陵沟壑区水土保持林为例[J].生态学杂志,2009,28(9):1737-1742.
[8]吴炳方.全国农情监测与估产的运行化遥感方法[J].地理学报,2000,55(1):23-35.
[9]李耀辉,王金鑫,李颖.应用时间序列EVI的MERSI多光谱混合像元分解[J].遥感学报,2016,20(3):459-467.
[10]马延辉,林辉,孙华.湖南省主要森林类型植被指数动态变化分析[J].水土保持研究,2010,17(3):217-221.
[11]王正兴,刘闯,ALFREDO H E A.植被指数研究进展:从AVHRR-NDVI到MODIS-EVI[J].生态学报,2003,23(5):979-987.
[12]赵英时.遥感应用分析原理与方法[M].北京:科学出版社,2003.
[13] GREEN R O,ROBERTS D A,GARDNER M,et al.Mapping chaparral in the Santa Monica Mountains using multiple spectral mixture models[J]. Remote Sensing of Environment,1998,65(3):267-279.
[14]李慧,张金区,曹阳,等.端元可变非线性混合像元分解模型[J].测绘学报,2016,45(1):80-86.
[15]蔡薇.基于MODIS遥感数据混合像元分解的小麦识别及其面积估算[D].济南:山东师范大学,2010.
[16]赵文吉,段福州,刘晓萌. ENVI遥感影像处理专题与实践[M].北京:中国环境科学出版社,2007.
[2] DYMOND J R,STEPHENS P R,NEWSOME P F,et al.Percentage vegetation cover of a degrading rangeland from SPOT[J]. International Journal of Remote Sensing,1992,13(11):1999-2007.
[3]王圆圆,李京.遥感影像土地利用/覆盖分类方法研究综述[J].遥感信息,2004,19(1):53-59.
[4]罗红霞,龚健雅.线性和非线性光谱混合模型模拟土壤、植被混合光谱的效果分析[J].测绘通报,2005(5):6-10.
[5]王天星,陈松林,马娅.基于改进线性光谱分离模型的植被覆盖度反演[J].地球信息科学,2008,10(1):114-120.
[6] RIDD M K. Exploring a V-I-S(vegetation-impervious surfacesoil)model for urban ecosystem analysis through remote sensing:comparative anatomy for cities?[J]. International Journal of Remote Sensing,1995,16(12):2165-2185.
[7]李巍岳,胡志斌,闫巧玲.基于ETM~+影像的森林资源信息提取———以黄土高原丘陵沟壑区水土保持林为例[J].生态学杂志,2009,28(9):1737-1742.
[8]吴炳方.全国农情监测与估产的运行化遥感方法[J].地理学报,2000,55(1):23-35.
[9]李耀辉,王金鑫,李颖.应用时间序列EVI的MERSI多光谱混合像元分解[J].遥感学报,2016,20(3):459-467.
[10]马延辉,林辉,孙华.湖南省主要森林类型植被指数动态变化分析[J].水土保持研究,2010,17(3):217-221.
[11]王正兴,刘闯,ALFREDO H E A.植被指数研究进展:从AVHRR-NDVI到MODIS-EVI[J].生态学报,2003,23(5):979-987.
[12]赵英时.遥感应用分析原理与方法[M].北京:科学出版社,2003.
[13] GREEN R O,ROBERTS D A,GARDNER M,et al.Mapping chaparral in the Santa Monica Mountains using multiple spectral mixture models[J]. Remote Sensing of Environment,1998,65(3):267-279.
[14]李慧,张金区,曹阳,等.端元可变非线性混合像元分解模型[J].测绘学报,2016,45(1):80-86.
[15]蔡薇.基于MODIS遥感数据混合像元分解的小麦识别及其面积估算[D].济南:山东师范大学,2010.
[16]赵文吉,段福州,刘晓萌. ENVI遥感影像处理专题与实践[M].北京:中国环境科学出版社,2007.