Identifying areas of mining Amethyst using ASTER images and digital image processing

Abstract

This work presents the identification of amethyst mining fields in the Ametista do Sul town (Rio Grande do Sul province) area using digital processing of satellite images from ASTER sensor. Remote sensing images for mapping and discrimination of geological targets is based on the interaction between electromagnetic radiation (EMR) and the targets. The radiation-target interaction is registered in terrain units, represented by the digital counter value of each pixel in the image. The observed values correspond to the measurement of a complex process of energy exchange involving EMR (transmission, reflexion and emission). This interaction occurs through atomic/molecular absorption/emission of EMR while interacting with atoms and molecules of the rock's minerals. Based on this concept, the manner in which these mineral compounds interact with EMR is identifiable by means of their spectral answers, that is, how each mineral transmits, reflects, absorbs and emits EMR. Cross-referencing this signature with the corresponding values allows statistical inference on which type of rock or mineral is present at a certain pixel of a digital image. With this in mind, we used: an ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) image from the Ametista do Sul region, duly calibrated and corrected; a group of spectral signatures corresponding to the main mineralization compounds and mining rejects of amethyst; and a posterior analysis of the correspondence between these signatures and the ASTER image values, obtained through DCDP recognition patterns (first derivative). Based on this method, thematic images were constructed to indicate the probability of amethyst mining, and were later validated with field data. This work demonstrates that the use of satellite data is an important tool in geological and field mining activity monitoring.