作　　者： ; ; ;

机构地区： 复旦大学信息科学与工程学院电子工程系

出　　处： 《复旦学报（自然科学版）》 2010年第2期197-208,I0003,共13页

摘　　要： 利用遥感图像进行城市不透水层分布的估算,在监测城市环境和规划城市发展等方面有着非常重要的作用.采用全约束的线性光谱混合模型,用3种方法(像元纯度指数、归一化的像元纯度指数和结合地表温度辅助分析的归一化像元纯度指数)分别从Landsat 7 TM/ETM+影像中分解得到上海市中心城区的不透水层分布,并比较了2002～2008年的上海市中心城区不透水层分布的变迁.结果表明,利用这3种方法分解得到不透水层的精度依次递增;总体来说,上海市不透水层分布比率较高.从2002～2008年不透水层分布变化来看,上海市的城市化有显著向外环线外区域扩张的趋势. Estimating the distribution of impervious surface using remote sensing images is very important in monitoring urban environments and programming urban developments. In present work, the impervious surface distribution is estimated through a fully constrained linear spectral mixture model using Landsat TM/ETM＋ data for the metropolitan area of Shanghai. Comparison of the accuracy of three methods, i. e. , pixel purity index （PPI）, normalized pixel purity index （NPPI）, and land surface temperature assistant normalized pixel purity index （T-NPPI） is carried out. Furthermore, the variation of the impervious surface distribution from year 2002 to year 2008 is analyzed. Experimental results show that T-NPPI leads to the best result while PPI is the worst one; it is shown that the ratio of impervious surface fraction to the total area of Shanghai is at a very high level. The spatial pattern of impervious surface of Shanghai is sprawling from the urban areas to the suburb areas by comparing the impervious surface distribution variation during the period of year 2002 to year 2008.

关 键 词： 不透水层 线性光谱混合分析 最小噪声分量 像元纯度指数 归一化 地表温度

领　　域： [自动化与计算机技术] [自动化与计算机技术]