pixelcooker

How to perform minerals identification and classification of rocks and soil using spectral angle mapper classifier and ENVI software. This demo applied using ASTER imagery, however it is can be extended to Landsat or Sentinel Imagery. Works better in open terrain such as arid or desert region with minimum to none of vegetation cover. Surface reflectance data (atmospherically corrected) is mandatory for the best result. Field inspection also mandatory to assess the classification accuracy because SAM is very sensitive to different parameter settings.

Color composites generation workflow which emphasizes to mineral (clay, iron oxide, and ferrous minerals) distribution over wide area. This workflow can be applied to Landsat-8 or Landsat-9 imagery. Similar workflow also can be done for Sentinel-2 MSI imagery. Workflow can be applied to Level-1 data, and the most optimal is applied to the surface reflectance level data. Key interpretation about color and mineralogy association: 1. Reddish color = ferrous minerals 2. Greenish color = clay minerals 3. Bluish color = iron oxide minerals