Indonesia Aplikasi Berbasis Web untuk Menentukan Kesegaran Daging Berdasarkan Gambar Menggunakan Algoritma Mobilenetv2

Elvino Junior, Syuhada Aqlul Hafiz, Muhammad Firaz Putra S.A, Fahrul Alamsyah, Agung Purnomo, Roby Saidi Prasetyo, Muhammad Firdaus, Mei Lestari, Ni Wayan Parwati SeptianiNon-Scopus21 Januari 2026Afiliasi: ID

Abstrak

The process of assessing meat freshness in the community still largely relies on subjective, manual methods, which can lead to errors and pose health risks to consumers.This study aims to develop a web-based application that objectively and accurately determines meat freshness from images.The method used is Research and Development (R&D), implementing the Convolutional Neural Network (CNN) algorithm within the MobileNetV2 architecture with the Transfer Learning technique.The image dataset was split into 80% for training, 10% for validation, and 10% for test.The model evaluation results show very satisfactory performance with an overall accuracy rate of 99%.In detail, the model achieved a precision, recall, and F1score of 0.98, 1.00, and 0.99 for the fresh meat category, and 1.00, 0.98, and 0.99 for the non-fresh category.These results prove that the developed application can be an effective solution for quick, accurate, and easily accessible meat quality verification.

Identifier

Ekspor BibTeX
Scroll to Top