BENTİK FORAMİNİFER GÖRÜNTÜ SINIFLAMASI VE TANIMLAMALARINDA EVRİŞİMLİ SİNİR AĞI (CNN) TABANLI YENİ BİR MODEL

Fossil studies are of great importance in order to observe the change living species over years, make inferences by using information provided observed species, and understand developing changing structure world we live years. However, examination interpretation fossil specimens is a complex long process. Artificial intelligence have begun be applied this field facilitate working methods paleontologists. The detection classification with aid computers simplifies process as much possible compared manual processes reduces foreign dependency for assemblages which paleontologists not experts. To achieve this, 9 benthic foraminiferal non-foraminiferal sample photographs from selected dataset were used. In study, new method developed foraminifera deep convolutional neural networks, reaching higher accuracy than results literature, presented. With method, at least 70% rates achieved test trained system. This reached high has created successful development branch paleontology use artificial microfossil identification.