The ground-based imaging hyperspectral remote sensing, having an advantage of integrated image and spectrum, is a frontier direction in the remote sensing field, and it can be applied to directly identifying surface objects based on spectral characteristics. This study took the imaging hyperspectral data for alteration belts in the Asiha gold ore district of Dulan county as an example, established a standard data processing workflow, in which the flat field method based on statistical models was applied to reconstruct spectrum, and the end member selection based on expert knowledge was utilized for mineral mapping using MTMF method. Based on above, true color images at a scale of 1∶100 and distribution of five hyperspectral surface objects (limonite, muscovite, etc.) were mapped. It is concluded that the typical alteration minerals in this ore district are limonite and muscovite, and ideal results are achieved after field verification using ASD spectral measurements. It is suggested that the ground-based imaging hyperspectral data have good application results and show a promising potential in geology.