多源数据的面向对象国际河流土地覆被分类研究

Object-oriented Landcover Classification of Multi-source Remote Sensing Data in International Trans-boundary River

作为横跨3个国家(尼泊尔、印度、中国)的国际跨界河流——柯西河流域,地形高差巨大,土地覆被结构组成复杂,进行土地覆被的自动分类研究具有典型意义。基于面向对象方法多源遥感数据、训练规则、丰富的细节信息为复杂土地覆被自动分类研究提供了可能。选择合适的影像分割特征和最优分割尺度,按照数据挖掘中的规则顺序逐步进行各个土地覆被的提取。总体精度说明分类结果与野外点相一致的概率能达到90.05%,说明国际跨界河流土地覆被分类方法是可行的,分类结果是准确、可信的。

Land cover classification is not easy for its inner complexity resulting from huge terrain elevation in Kosi River as international trans-boundary river, it has possible to classification difficulties flowing through three countries. The emergency of multi-source remote sensing images and training algorithm make it possible to classify the land cover for its ample details based on object-oriented method. The paper is about land cover classification method selecting feature and optimal scale of segmentation, the innovation of this method lies in selection of proper scale parameter resulting from proper image data and certain classification order. The total accuracy has highly 90.05% compared with object oriented classification result and actual sampling point, which is a feasible method, and the classification results are more accurate.