Emotion Recognition System for Game Addiction Users Using FACS and Image Feature Extraction on Android

Roslinda Tanjung (Universitas Muhammadiyah Sumatera Utara)
Yohanni Syahra (Universitas Muhammadiyah Sumatera Utara)

Tsabit
Vol. 3 No. 1 (2026): June Edition

Purpose: Online game addiction negatively impacts users’ mental health and social behavior. This study aims to develop an Android-based emotion recognition system using Facial Action Coding System (FACS) with image feature extraction to detect users’ emotions in real-time. Design/Methods/Approach: The system integrates OpenCV for face detection, dlib for facial landmark identification, and TensorFlow Lite for emotion classification. Testing was conducted through black box testing, confusion matrix evaluation, and performance analysis. Findings/Results: The system successfully recognized basic negative emotions (anger, frustration, sadness, displeasure) with adequate accuracy (87%) and operated effectively on Android devices. Conclusions: Integrating FACS with image feature extraction provides a non-intrusive solution to help users recognize and manage emotions, contributing to digital mental health research and practice.