Abstract: Various hyperspectral analysis methods proposed in recent years have been successfully applied to the extraction of altered minerals, but the effects of different extraction methods vary differently. As information of altered minerals in Northwest China tends to be weak and hard to identify, how to quantitatively compare the effectiveness of different analysis methods is worth further investigation. Taking the CASI-SASI airborne hyperspectral data of Laojinchang in Beishan area as an example, the “minimum noise separation-pure pixel index-n-dimensional visualization” technology was used to analyze the key thresholds in the dimensionality reduction and denoising of hyperspectral data and the extraction of pure pixels. The authors quantitatively compared the effectiveness of five extraction methods (MTMF, SFF, SID, SAM and ACE) with the help of the similarity increment and spectral distinction entropy index. The ACE method, with a relatively better extraction effect, was selected to extract altered mineral information in the experimental area. The results were finally tested by means of typical geological section verification. The results show that the MNF transform can effectively sort the bands according to the image quality to achieve data dimensionality reduction and denoising; the optimal algorithm threshold of the PPI appears when the iterative curve becomes close to the horizontal line. Among the methods mentioned, the ACE and MTMF methods are more sensitive to slight changes in the spectral curve and the comparison indexes of the two methods are basically the same. Field verification of the geological section found out that the types of altered minerals extracted in experiments was consistent with the types of surface altered geological bodies, and the surrounding rock alteration of the deposit was effectively reflected by the altered mineral information combination. The research provides theoretical basis and technical reference for the extraction and application of altered mineral information in Gobi desert areas of Northwest China.