Evaluation and Comparison of PCA Processing Results for ASTER Data in SWIR and VNIR + SWIR Ranges (Case Study: Firouzabad Salt Domes, Fars)

Nowadays, different methods of processing satellite imagery are used to distinguish different rocks and minerals. The ASTER sensor due to its higher bands than older sensors, its desirable spatial resolution and low-cost data has attracted the attention of many researchers and various processing methods can be applied on its data. The reflection range of this sensor has the greatest potential for investigating the composition of natural materials on earth and the absorption spectrum of hydroxyl, sulfate and carbonate minerals is in this range. In this study, Principal Component Analysis (PCA) was performed in the reflection (VNIR + SWIR) and 6-band (SWIR) ranges for these data. Then, in order to evaluate and compare the results of PCA processing, false color composite of the output components with the highest eigenvalue in the SWIR and VNIR+SWIR ranges was created which resulted appearing different areas on the Firouzabad salt domes, Fars (Konarsiah & Jahani) with different colors. The results were compared with geological map, field observations and the accuracy of processing was evaluated.