Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : METHOMIKA: Jurnal Manajemen Informatika

Sistem Pendeteksi Tingkat Kesegaran Daging Ayam pada Citra Menggunakan Metode Convolutional Neural Network (CNN) Berbasis Android Naturizal, Rayhan; Fuadi, Wahyu; Rosnita, Lidya
METHOMIKA: Jurnal Manajemen Informatika & Komputerisasi Akuntansi Vol. 8 No. 2 (2024): METHOMIKA: Jurnal Manajemen Informatika & Komputersisasi Akuntansi
Publisher : Universitas Methodist Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46880/jmika.Vol8No2.pp301-312

Abstract

This research develops a chicken meat freshness detection system based on image processing, implemented on an Android platform using the Convolutional Neural Network (CNN) method optimized with TensorFlow Lite. The system classifies chicken meat into three categories: fresh, less fresh, and rotten. The CNN model uses 32 filters to enhance feature extraction from the meat images. Testing on 30 samples, with each category tested 10 times, showed an accuracy of 90%, with 27 correct detections and 3 errors in the less fresh category. While the system effectively identifies fresh and rotten categories, there is a challenge in distinguishing the less fresh category due to its ambiguous visual characteristics. One limitation is the lack of a bounding box, causing the application to still provide detection results even when the scanned object is not chicken meat. This application is specifically designed to detect chicken meat pieces, so it is not recommended for use outside this context.