Endemic Plant Classification Using Deep Neural Networks

Abstract

Endemic plants are those that are native to a specific geographic region and are found nowhere else in the world. These plants are crucial for biodiversity, conservation, cultural significance, and economic value. Turkey hosts more than 4000 endemic plants. Therefore, this makes Turkey the richest in Europe. Preserving this habitat holds importance. This study aims to conceptualize a possible application that helps individuals to identify endemic species using camera-captured images. Thus, aiding the preservation of the habitat. In this study, 23 selected species of Turkey’s endemic biodiversity are classified using Deep Neural Network built. In line with the objective of this study, a dataset containing 253 images is created to train the network. The dataset is available at: github.com/melihoz/endemicdataset