中文摘要：

为了对比CBERS与TM两种遥感影像在地表覆被信息提取中的具体性能,验证基于CBERS遥感影像进行湿地覆被分类的可行性,以典型内陆淡水湿地区为对象,基于CBERS与TM遥感影像,针对各波段进行信息量统计及光谱特性分析,获取了各波段覆被探测性能的初步认识;运用非监督、监督与面向对象三种代表性分类方法进行分类实验,通过精度误差矩阵对比分类结果,分析了两种遥感影像在湿地覆被分类中的准确程度差异;基于分类结果,通过景观格局指数计算,对比分析了两种影像在湿地覆被信息提取结果上的空间差异和特性。

英文摘要：

Taking the typical inland wetland as object,based on CBERS and TM remote sensing images,through the information quantity statistics and spectral characteristic analysis,we obtain the cognition ability for each band of landcover.After classification using unsupervised,supervised and object-oriented methods,and through the analysis of error matrix,the accuracy difference between two images is obtained.Based on the classification result,through the calculation of landscape index,the spatial difference and characteristics of wetland cover classification results are analyzed.