顾及植被的复杂艰险地区多光谱遥感岩性识别

Lithology identification of multi-spectral remote sensing in complex and difficult areas considering vegetation coverage

针对复杂艰险地区修建道路时,地质岩性难以现场获取的问题,该文利用ASTER多光谱遥感影像,研究和设计了一种顾及植被覆盖的复杂艰险地区多光谱遥感岩性信息识别提取方法。该方法根据植被覆盖情况先分类后进行岩性识别:对中高植被覆盖区进行植被信息的抑制处理,再对中低植被覆盖区采用主成分分析法进行岩性信息提取;对低植被地表裸露区域采用端元提取、光谱角填图的方法进行岩性信息的提取。最后将岩性识别提取结果与现场地勘结果进行对比验证。结果表明该方法进行岩性信息提取结果与已有的现场勘察结果基本一致,具有重要的参考和应用价值。

Aiming at the problem that geological lithology is difficult to obtain on site, when constructing roads in complex and difficult areas, in this paper, using advanced spaceborne thermal emission and reflection radiometer(ASTER) multi-spectral remote sensing images, we studied and designed a multi-spectral remote sensing lithological information identification and extraction method that took into account the complex and difficult areas of vegetation coverage. The method was to classify and then perform lithology identification according to the vegetation coverage: the vegetation information was suppressed in the middle and high vegetation coverage area, and then the lithological information was extracted by the principal component analysis method in the middle and low vegetation coverage area;lithological information was extracted by end-element extraction and spectral angle mapping. Finally, the lithology identification and extraction results were compared and verified with the on-site geological survey results. The results showed that the lithological information extraction results of this method were basically consistent with the existing on-site survey results, and had important reference and application value.