Claim Missing Document
Check
Articles

Found 2 Documents
Search

Performance Evaluation of Food Calorie Counter Mobile Application Based on CNN-YOLO Algorithms Hamzidah, Nurul Khaerani; Ulandari, Ayu; Parenreng, Mardawia Mabe; Ichzan As, Nur
Jambura Journal of Electrical and Electronics Engineering Vol 7, No 2 (2025): Juli - Desember 2025
Publisher : Electrical Engineering Department Faculty of Engineering State University of Gorontalo

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37905/jjeee.v7i2.30595

Abstract

This article discusses the performance results of a mobile application for calculating food calories. This application can help users in managing and knowing their daily calorie intake and a healthy diet program. This application is based on image processing techniques using a combination of CNN-YOLOv5. The role of CNN is to classify data and extract labeled food image features using the supervised learning method based on images that have gone through the training process stages while YOLO plays a role in detecting food image data quickly and accurately. The design stages consist of UI design, UI creation, program implementation, and testing and evaluation. In analyzing the model, 1,736 training data, 149 test data and 206 validation data were used with 150 epochs of computation. The results of the model analysis obtained a precision of 1.00, confidence 0.962, recall 0.99 and F1 score 0.97. These results indicate that the system has met the requirements for use in further detection processes. This is evidenced by the application's ability to detect food images with 100% accuracy for all food classes in real-time or through image uploads. The test results show that the confidence value is influenced by the distance of the detector to the object, lighting intensity, camera resolution, color similarity, food variety and the background motif of the container used. The application is equipped with attractive features and UI displays such as an informative BMI calculator especially for users who are on a healthy diet program. The application of the CNN-YOLOv5 algorithm combination has been proven to be able to consistently and accurately improve application performance in detecting types of food and their calorie content in 100 grams so that it is worthy of being used as a reference in helping to monitor daily calorie intake and help a healthy diet program.Artikel ini membahas hasil kinerja aplikasi mobile penghtiung kalori makanan. Aplikasi ini dapat membantu pengguna dalam mengatur dan mengetahui asupan kalori harian serta program diet sehat. Aplikasi ini berbasis teknik pengolahan citra menggunakan kombinasi CNN-YOLOv5. Adapun peran CNN adalah untuk mengklasifikasi data serta mengekstraksi fitur citra makanan yang telah terlabel dengan menggunakan metode supervised learning berdasarkan citra yang telah melalui tahapan proses training sedangkan YOLO berperan dalam mendeteksi data citra makanan dengan cepat dan tepat. Tahapan perancangannya terdiri dari perancangan desaian UI, pembuatan UI, implementasi program, dan pengujian serta evaluasinya. Dalam menganalisis model digunakan 1.736 data latih, 149 data uji dan 206 data validasi dengan komputasi 150 epoch. Hasil analisis model diperoleh presisi 1.00, confidence 0.962, recall 0.99 serta F1 score 0.97. Hasil ini menunjukkan bahwa sistem sudah memenuhi syarat untuk digunakan dalam proses deteksi lebih lanjut. Hal ini dibuktikan dengan kemampuan aplikasi dalam mendeteksi citra makanan dengan akurasi 100% untuk semua kelas makanan secara real-time ataupun melalui upload citra. Hasil pengujian menunjukkan bahwa nilai confidence dipengaruhi oleh faktor jarak detektor ke objek, intensitas pencahayaan, resolusi kamera, kemiripan warna, variasi makanan serta adanya motif background wadah yang digunakan. Aplikasi dilengkapi dengan fitur dan tampilan UI yang menarik seperti kalkulator BMI yang informatif khususnya bagi pengguna yang sedang dalam program diet sehat. Penerapan kombinasi algoritma CNN-YOLOv5 terbukti mampu meningkatkan kinerja aplikasi secara konsisten dan akurat dalam mendeteksi jenis makanan beserta kandungan kolari dalam 100 gramnya sehingga layak dijadikan sebagai rujukan dalam membantu monitoring asupan kalori harian dan membantu program diet sehat.  
AI-YOLO Based Smart Laboratory Security for Automated Face Recognition and Suspicious Activity Detection Hamzidah, Nurul Khaerani; Syahrir, Syahrir; Jariyah, Ainun; Da Costa, Carlos Agunar; Saenab, Sitti; Muin, Dul Arafat; Ichzan As, Nur
Journal of Applied Informatics and Computing Vol. 10 No. 1 (2026): February 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i1.11936

Abstract

Ensuring laboratory security is a critical consideration within campus environments to effectively prevent theft and suspicious activities. Traditional CCTV systems predominantly rely on manual monitoring, resulting in delayed responses to incidents. This research seeks to develop and implement an Artificial Intelligence (AI)-based laboratory security system, integrating three primary models: YOLOv5 for human object detection, Face Recognition for individual identification, and Media Pipe Pose for real-time analysis of suspicious movements. The system is designed as a Flask-based monitoring website, which displays activity logs, detected individual identities, and automated notifications based on image processing results on a Raspberry Pi connected to CCTV cameras. The research methodology employs an applied experimental approach, encompassing stages such as system design, face dataset collection, data encoding utilizing the Face Recognition Library, and performance evaluation under two lighting conditions (bright and dark) and three distance variations. The test results indicate that the Face Recognition method operates optimally at a distance of 2 meters in bright lighting conditions, achieving an accuracy of up to 92%. However, its performance declines at distances exceeding 3 meters and under low-light conditions. Conversely, MediaPipe Pose exhibits high stability, with an average accuracy of 94% in bright conditions and 89% in dark conditions, and is capable of transmitting real-time notifications for activities such as lifting objects or placing hands into pockets. The AI-based laboratory security system developed has demonstrated effectiveness, adaptability, and responsiveness in the automatic detection of identities and suspicious activities. The integration of YOLO v5, Face Recognition, and MediaPipe Pose models offers an intelligent and efficient security solution that facilitates the implementation of the Smart Campus concept within educational environments.