【摘要】 为提高高分辨率遥感影像的分类精度 ,本文提出了一种 GIS与 RS集成的分类技术。它从遥感影像和 GIS矢量数据一体化的角度出发 ,充分利用了矢量数据的图斑边界信息 ,通过提取单一地类图斑内的灰度特征、纹理特征和形态特征识别图斑所属地类。经研究表明 :无论在实验结果上还是在分类的机理上都证明了 ,在高分辨率遥感影像的土地利用分类中 ,这种 GIS与 RS集成的分类技术的准确率超过了传统遥感影像分类的准确率更多还原

【Abstract】 In order to improve the classifying precision of high resolution images, this paper presented a specific kind of GIS & RS integration technology. It can make full use of GIS vector information (Polygon Borderline), and could identify any test parcel’s land type by comparing with standard land type in terms of abstracted gray character, texture character, and shape character within borderline of polygon. The result of practical tests and the mechanism analysis all verified the higher nicety degree of such kind of method over the traditional method.更多还原