基于混合像元分解提取多种类覆盖区遥感蚀变信息 ——以甘肃省岷县寨上金矿区为例

在多种类覆盖区，由于植被、河道、阴影、第四系等干扰因素影响，难于客观有效提取与成矿有关的遥感蚀变信息。应用混合像元分解的方法技术可以区分出混合像元中各类干扰信息的反射率值及其所占丰度值，从混合像元光谱中减去各干扰因素的反射率值及其所占份额，可以获取土壤或岩石的反射率值及其所占丰度值，将土壤或岩石反射率值除以其所占的份额...

Extracting Remote Sensing Alteration Information from areas with Various Covers Based on Decomposition of Mixed Pixels: An Example of the Zhaishang Gold Mine in Min County of Gansu Province

Abstract:It is difficult to extract remote sensing alteration information related with mineralization due to the influences of interference factors such as vegetation, river courses, shadows, and Quaternary sediments in areas with various covers. The reflectivity data and its abundance of various disturbance information in mixed pixels can be distinguished using the decomposition technology of mixed pixels. Subtracting the reflectivity values and their occupying ratios each interference factor from spectra of mixed pixels,the reflectivity values and their abundance of soil or rocks can be determined. The spectra of soil or rocks are reestablished by dividing the reflectivity of the soil or rock with its portion. The reliability of extracting remote sensing alteration information can be improved based on the new images after the effects of interference factors are eliminated, even using the traditional principal component analysis method. This technology has been applied to the Zhaishang gold mine in Min County of Gansu Province and satisfactory results have been achieved.