Research on information extraction of iron tailings from multi-source remote sensing images coupled with hyperspectral data

Tailings are the main waste in the process of mining and beneficiation of mineral resources, and the treatment method is mainly stacking and discharge, which pollutes the environment and is not conducive to resource recycling.Using the high-efficiency and convenient technical advantages of remote sensing to comprehensively find out the stacking status of tailings is an important help to promote the realization of the goal of “double carbon”.Taking Sijiaying Tailings Pond in Luan County, Tangshan City, Hebei Province as the research object, the measured tailings spectrum is used to identify the end element spectrum through spectral angle match (SAM) and spectral feature fitting (SFF), and then the tailings are mapped based on Landsat8 OLI image and Zhuhai-1 image data.The results show that the matching score of tailings in Landsat8 OLI image is 1.618, and the matching score of tailings in Zhuhai-1 image is 2.069.The recognition results of multispectral data and hyperspectral data are highly consistent.The mapping accuracy of Zhuhai-1 hyperspectral image is better than that of Landsat8 OLI multispectral image.The mapping results can provide data reference for mineral resources management departments.