论文部分内容阅读
专题指数对遥感影像自动解译至关重要,现有研究多针对单目标信息提取来筛选专题指数,无法得到适用于多目标遥感自动解译的最佳专题指数。以德州市城区及周边地区为例,采用Landsat 5TM影像提取了2个植被、3个水体和3个建筑用地专题指数,基于面向对象分类方法,分析了单个专题指数、指数组合、指数数量对同时提取植被、水体和不透水层信息的精度影响。结果表明:(1)3类地物的最小分类精度基本上随着专题指数增加而增大;(2)从单个专题指数来看,不透水层和植被提取的最佳指数分别是建筑物指数和土壤调整植被指数,而新型水体指数则能显著提高总体分类精度;(3)从专题指数的组合来看,植被分类精度随所用的植被指数数量增加而下降;建筑用地指数越多,不透水层和总体分类效果越好;随着水体指数数量增加,水体分类精度有所提高,而不透水层和总体分类精度则随之下降。
The thematic index is very important for the automatic interpretation of remote sensing images. In the existing research, the thematic index is screened for single-target information extraction, and the best thematic index suitable for multi-target remote sensing automatic interpretation can not be obtained. Taking the urban area and the surrounding area of Dezhou as an example, two thematic indexes of vegetation, three bodies of water and three construction sites were extracted using the Landsat 5 TM imagery. Based on the object-oriented classification method, a single thematic index, index combination, Accuracy of information extracted from vegetation, water and impermeable layers. The results show that: (1) The minimum classification accuracies of three kinds of land cover basically increase with the increase of the thematic index; (2) From the single thematic index, the best index of impermeable layer and vegetation extraction are the building index And soil adjustment vegetation index, while the new water body index can significantly improve the overall classification accuracy; (3) From the combination of thematic indices, the accuracy of vegetation classification decreases with the increase of the vegetation index; the more the index of construction land, the impervious The better the classification of the water layer and the overall classification, the higher the classification accuracy of the water body with the increase of the index of the water body, while the impermeable layer and the overall classification accuracy will decrease.