基于地物光谱分析的ASTER岩土类型识别

Identification of soil types from ASTER image based on spectral analysis

利用遥感数据和实测不同岩土类型波谱测试数据,以实测不同土体反射光谱数据为基础,通过对遥感数据进行图像增强处理,对不同沉积环境、不同粒级组成的松散土体类型进行识别。研究发现,ASTER数据721波段组合能够有效识别不同土体类型,在凸显不同土体类别的效果最好;ASTER数据的平方根拉伸及标准分段拉伸对不同土体类型的区分效果最好;通过对遥感数据进行主成分变换,发现遥感数据图像信息量得到有效集中,增强了图像的层次感,能够有效增强图像信息。

By making use of enhanced remote sensing data and on the basis of measured reflectance spectrum of different geotechnical species, the authors achieved type identification of loose soils which composed of different grain sizes in different depositional environments. The study found that, ASTER data wkh band combination of 721 could effectively identify the different geotechnical species with the best effect in highlighting different geotechnical species. Besides, ASTER image enhancement by square root stretching, standard segment stretching and principal component analysis application could effectively enhance the image information and improve the identification accuracy.