Rahimzadegan, Majid, Sadeghi, Behnam, Masoumi, Manuchehr, and Ghalehjoghi, Siamak Taghizadeh (2015) Application of target detection algorithms to identification of iron oxides using ASTER images: a case study in the North of Semnan province, Iran. Arabian Journal of Geosciences, 8 (9). pp. 7321-7331.

PDF (Published Version) - Published Version Restricted to Repository staff only

 

12

2

Abstract

In recent years, satellite sensor images have been commonly used in mineral detection. These images, with different spectral and spatial resolutions, can be used in detecting minerals which have surface indicators. Due to numerous sources of uncertainties in using spectral processes for mineral explorations and atmospheric effects on the spectrum of the pixels, in this research, mineral target detection methods were selected and implemented. The implementation of this research was done using an ASTER scene from the northern area of Semnan province in Iran. There were seven iron mines in the region covered by this scene. Three of them were used as training data and four other ones were used as test data. Several target detection methods were implemented and the mean of the results of these algorithms, which embrace the results of all algorithms, was evaluated. The output of the algorithm is an image where the gray value of each pixel corresponds to the probability of similarity to the training data. Considering the fact that the probabilities' range is between 0 and 1, after implementing the algorithms, it was concluded that the spectral angle mapper (SAM) method has the best performance with mean probability value of 0.99 for the test mines. Based upon the fact that the mean value of the algorithms was 0.87, it was proved that these methods can be very practical thanks to their high accuracy in prospection of new exploration targets which may detect new equivalent iron potentials.

Actions (Repository Staff Only)

Item Control Page