Mineral type remote sensing recognition method based on multi-type spectral feature parameter collaboration

Abstract

The invention discloses a mineral type remote sensing recognition method based on multi-type spectral feature parameter collaboration. The method includes the steps of: a. conducting spectral re-building on hyperspectral data; b. for different hyperspectral data, conducting spectrum resampling on typical mineral spectral data according to different sampling intervals; c. firstly conducting envelope removal treatment on hyperspectral data and a typical mineral spectrum, and then extracting multi-type spectral feature parameters respectively; d. calculating the information amount of different spectral feature parameter combinations through an optimum index factor, determining an optimum spectral feature parameter combination, and conducting mineral mapping experiment based on pattern recognition method; and e. firstly carrying out statistical analysis on the multi-type spectral feature parameters of the typical mineral spectrum, then referring to spectral feature parameter values corresponding to mineral end members in the hyperspectral data, constructing a research area mineral type recognition decision tree model, and conducting mineral mapping experiment. The mineral type remote sensing recognition method provided by the invention has high integral mineral recognition precision.

Description

technical field [0001] The invention relates to a mineral type remote sensing identification method based on the coordination of multi-type spectral characteristic parameters, which is suitable for the field of regional mineral mapping of various hyperspectral data. Background technique [0002] Because hyperspectral remote sensing data has the characteristics of continuous band and high spectral resolution, it can play an important role in the identification of mineral types, and has been gradually and widely used in local area mineral mapping and other work. The use of remote sensing means can realize the rapid identification and extraction of mineral information from a wide-area space and a multi-temporal scale, shorten the time of mineral mapping, and improve the efficiency. However, due to the limitation of complex surface structure, mixed pixels and atmospheric environment background, the accuracy of mineral identification using remote sensing technology is not high. ...

AI Technical Summary

Problems solved by technology

In areas with a single mineral type, this method can show high accuracy, but in areas with complex mineral types, since the spectral characteristics of minerals are mainly concentrated in the very narrow wavelength region of short-wave infrared, it is difficult to use a single spectral characteristic parameter Differentiate between different types of minerals

Therefore, the limitations of current technical methods limit the application of hyperspectral remote sensing technology in mineral type identification

Images