基于高分影像纹理分维变化的灾害自动识别方法
|
摘要
目前遥感变化检测主要是人机交互式目视解译,不能满足灾害自动识别尤其是卫星在轨变化检测自动化的新要求,亟须研究遥感影像变化自动检测算法,以实现灾害遥感自动识别。该文将高空间分辨率遥感影像的多尺度格网分割与纹理分维变化比较相结合,提出基于高分影像纹理分维单调变化(Texture Fractal MonotonousChange,TFMC)的灾害自动识别方法。通过计算和对比不同格网分割尺度下前后两期高分影像的纹理分维变化及其空间分布,并基于纹理分维变化单调性准则,可自动检测并识别灾区范围。以2011年3月11日日本地震海啸灾区的Worldview 0.5m全色影像为例,进行实验研究,表明MTFC方法无需人工干预即可根据纹理分形单调下降(当前减先前)可靠地识别出海水淹没区和密集房屋损毁区。经进一步优化,MTFC方法可望发展为高分遥感卫星在轨变化检测及灾害链聚焦监测的新技术。
Currently,the main techniques of change detection from remote sensing images is visual interpretation based on man-machine interaction,which cannot fit for the demands of disaster auto-recognition especially of change detection by orbiting satellites.Combined with the image partition by multi-scale grid and the comparison on texture fractal change of remote sensing images with High Spatial Resolution(HSR),and based on the rule of Texture Fractal Monotonous Change(TFMC),this paper presented a new method,named TFMC-method,to recognize natural disasters automatically.By calculating and comparing the texture fractal changes of HSR remote sensing images(current and before),the changed texture fractal and its spatial distribution inside the imaging zone are obtained,and possible disasters and its spatial distributions can be detected and recognized automatically.With Ms.9.0 Japan Earthquake and Tsunami,Mar.11,2011,being a case,the MTFC-method was experimentally tested and analyzed with WorldView 0.5 m panchromatic images.It was proved that the MTFC-method is efficient to recognize with high reliability the sea-water flooded zone and the damaged building zones referring to the monotonically decreased texture fractal,and no artificial work is needed.With further improvement,the MTFC-method is expected to be developed to be a new technology for orbiting satellite to detect earth surface disaster and to conduct focusing monitoring on disaster chains.
Currently,the main techniques of change detection from remote sensing images is visual interpretation based on man-machine interaction,which cannot fit for the demands of disaster auto-recognition especially of change detection by orbiting satellites.Combined with the image partition by multi-scale grid and the comparison on texture fractal change of remote sensing images with High Spatial Resolution(HSR),and based on the rule of Texture Fractal Monotonous Change(TFMC),this paper presented a new method,named TFMC-method,to recognize natural disasters automatically.By calculating and comparing the texture fractal changes of HSR remote sensing images(current and before),the changed texture fractal and its spatial distribution inside the imaging zone are obtained,and possible disasters and its spatial distributions can be detected and recognized automatically.With Ms.9.0 Japan Earthquake and Tsunami,Mar.11,2011,being a case,the MTFC-method was experimentally tested and analyzed with WorldView 0.5 m panchromatic images.It was proved that the MTFC-method is efficient to recognize with high reliability the sea-water flooded zone and the damaged building zones referring to the monotonically decreased texture fractal,and no artificial work is needed.With further improvement,the MTFC-method is expected to be developed to be a new technology for orbiting satellite to detect earth surface disaster and to conduct focusing monitoring on disaster chains.
引文
[1]陈爱群.国内外遥感最新技术及其发展趋势[J].测绘科技情报,2008(1):12-15.
[2]赵英时.遥感应用分析原理与方法[M].北京:科学出版社,2003.241-242.
[3]张继贤,杨贵军.单一时相遥感数据土地利用与覆盖变化自动检测方法[J].遥感学报,2005,9(3):294-299.
[4]孙晓霞,张继贤,燕琴,等.遥感影像变化检测方法综述及展望[J].遥感信息,2011(1):119-123.
[5]MANDELBROT B B.How long is the coast of Britain?Statisti-cal self-similarity and fractional dimension[J].Science,1967,156(3775):636-638.
[6]MANDELBROT B B.Stochastic models for the Earth′s relief,the shape and the fractal dimension of the coastlines,and thenumber-area rule for islands[P].Proc Natl Acad Sci U.S.A,1975.
[7]MANDELBROT B B.Self-affine fractals and fractal dimension[J].Physica Scripta,1985,32(4):257-260.
[8]栾海军,汪小钦.多源遥感影像分形特征分析[J].遥感信息,2010(3):7-12.
[9]张华国,黄韦艮,周长宝,等.关于IKONOS卫星遥感图像的分形特征研究[J].测绘通报,2005(5):15-18.
[10]杨彦从,彭瑞东,周宏伟.三维空间数字图像的分形维数计算方法[J].中国矿业大学学报,2009,38(2):251-258.
[11]陈文凯,何少林,周中红,等.地震前后遥感影像分形特征研究[J].大地测量与地球动力学,2010,30(6):24-30.
[12]CLARKE K C.Computation of the fractal dimension of topo-graphic surfaces using the triangular prism surface area meth-od[J].Computers&Geosciences,1986,12(5):713-722.
[13]SUN W,XU G.Fractal analysis of remotely sensed images:Areview of methods and applications[J].Int.J.Remote Sensing,2006,27(22):4963-4990.
[14]陈颙,陈凌.分形几何学[M].北京:地震出版社,2005.
[2]赵英时.遥感应用分析原理与方法[M].北京:科学出版社,2003.241-242.
[3]张继贤,杨贵军.单一时相遥感数据土地利用与覆盖变化自动检测方法[J].遥感学报,2005,9(3):294-299.
[4]孙晓霞,张继贤,燕琴,等.遥感影像变化检测方法综述及展望[J].遥感信息,2011(1):119-123.
[5]MANDELBROT B B.How long is the coast of Britain?Statisti-cal self-similarity and fractional dimension[J].Science,1967,156(3775):636-638.
[6]MANDELBROT B B.Stochastic models for the Earth′s relief,the shape and the fractal dimension of the coastlines,and thenumber-area rule for islands[P].Proc Natl Acad Sci U.S.A,1975.
[7]MANDELBROT B B.Self-affine fractals and fractal dimension[J].Physica Scripta,1985,32(4):257-260.
[8]栾海军,汪小钦.多源遥感影像分形特征分析[J].遥感信息,2010(3):7-12.
[9]张华国,黄韦艮,周长宝,等.关于IKONOS卫星遥感图像的分形特征研究[J].测绘通报,2005(5):15-18.
[10]杨彦从,彭瑞东,周宏伟.三维空间数字图像的分形维数计算方法[J].中国矿业大学学报,2009,38(2):251-258.
[11]陈文凯,何少林,周中红,等.地震前后遥感影像分形特征研究[J].大地测量与地球动力学,2010,30(6):24-30.
[12]CLARKE K C.Computation of the fractal dimension of topo-graphic surfaces using the triangular prism surface area meth-od[J].Computers&Geosciences,1986,12(5):713-722.
[13]SUN W,XU G.Fractal analysis of remotely sensed images:Areview of methods and applications[J].Int.J.Remote Sensing,2006,27(22):4963-4990.
[14]陈颙,陈凌.分形几何学[M].北京:地震出版社,2005.
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