Akademik Veri Yönetim Sistemi
Journal of Scientific and Engineering Research,, cilt.11, sa.9, ss.1-8, 2022 (Hakemli Dergi)
Abstract In geophysical studies, it is an important issue to determine the building boundaries and to perform
separation analysis. In addition to the classical techniques used in geophysical methods, the use of Artificial
Intelligence-based image processing techniques is considered attractive. Recently, Cellular Artificial Neural
Networks (CNN) method has been used in geophysical studies and very good results have been obtained. The
filtering structure of the CNN method provides fast and parallel computation capability for geophysical image
processing applications. The behaviour of the CNN method is adjusted by the supervised learning algorithm and
is defined by two pattern matrices. For the analysis of data with potential origin, the training phase is first
applied to the CNN method, processed into potential origin anomaly maps, and can be analysed sequentially. In
this study, the CNN method was applied to the vertical magnetic anomaly map made in the iron mine in the
Bingöl region of Turkey. CNN outputs were compared with the boreholes drilled later in the region. As a result,
it has been shown that the CNN method is a good method that can be used in solving the problems of both
separation and boundary detection.
Keywords Magnetic anomaly map, iron ore, Cellular Neural Network (CNN), Turkey-Bingö