论文部分内容阅读
以杂多县冰川为研究对象,为精确提取冰川信息,根据冰川在TM影像上存在的光谱差异,将冰川划分成两类。分析与周边地物较难区分的冰舌和其他地物之间的光谱特征区分性,选择最优的冰川提取波段;同时,利用波段运算得到对冰川信息提取更有利的波段,将所有优选波段经过最佳指数计算和地物可分离性分析,得到波段间相关性最低,冗余信息最少,冰川与其他地物可分性最强的波段组合进行监督分类,有效地分离出冰川。将该方法提取的冰川面积与高分辨率影像中提取的冰川面积相比较,结果表明该技术方法是有效的。
Taking the Heterodoxian glacier as the research object, in order to accurately extract glacier information, the glaciers are divided into two types according to the spectral difference existing in the TM images of the glacier. Glaucoma and other landforms that are more difficult to differentiate from the surrounding objects are analyzed to select the optimal glacial extraction band. At the same time, a more favorable band for glacier information extraction is obtained by using the band operation, and all the preferred bands After the best index calculation and the separability analysis, the combination of the bands with the lowest correlation, the least redundant information and the most separable glacier and other features can be obtained, and the glacier can be effectively separated. Comparing the area of glacier extracted by this method with the area of glacier extracted from high-resolution images, the results show that this method is effective.