Klasifikasi Jenis Ikan Hias Air Laut dengan Menggunakan Metode VGG-19 Network

Abstract

The various types of seawater ornamental fish in Indonesia with characteristics that are similar to each other make it difficult to identify seawater ornamental fish species directly by eye. The distinction or classification of seawater ornamental fish is currently still done manually and is assumed to be personal which makes the information presented less accurate. For this reason, need a system that can facilitate the process of classifying seawater ornamental fish species where the aim is to recognize them more effectively and also more efficiently. There are three species classified in this study, namely lori anthias, randali anthias, and smithvanizi anthias. By using the VGG-19 Network and also 450 data which is divided into 360 training data images and 90 images for testing. The research has been quite good based on the classification results obtained with an accuracy of 96,6%.