Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Jurnal Infra
Michael Christiawan
Program Studi Informatika
Computer Science & IT

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Deteksi Tingkat Kesegaran Ikan Menggunakan Metode Convolutional Neural Network Dengan Parameter Mata dan Warna Insang Michael Christiawan; Leo Willyanto Santoso; Djoni Haryadi Setiabudi
Jurnal Infra Vol 9, No 2 (2021)
Publisher : Jurnal Infra

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

The sea resources of Indonesia are very abundant. According to Indonesia Central Bureau of Statistics, in 2017, there are 16.114.991 tons of farmed fish production [2] and 7.361.116 tons of catch fish production in 2018 [3]. But sometimes when it’s already delivered, the freshness level is already decreasing, so we need a tool or an app to determine the level of fish freshness. Navotas, et al. [6] research use Artificial Neural Network to determine the fish freshness from 3 kind of fish. They got more than 90% accuracy. But they use a tool to take fish eye and gill picture. They also not classifying what kind of fish that is being detected.In this research, Convolutional Neural Network (CNN) is used for fish and freshness classification from 4 kind of fish. Detectron2 model is used to detect the fish eye position. Masking is used for detecting the gill. The picture taken with smartphone camera only.The CNN model is already succeed in classifying the fresh and stale fish eye, except for the grouper fish, so the accuracy is 75%. But the CNN model isn't good enough to classify the fresh and stale gills, because there is only a slight difference between the fresh, quite fresh, and stale gills. The accuracy for gill freshness model is 25%.
Co-Authors Djoni Haryadi Setiabudi Leo Willyanto Santoso