Deep Learning-Based Implementation of Convolutional Neural Networks for Skin Disease Detection Through Image Classification on Mobile Platforms

Nuzul Hikmah (Electrical Engineering Department, Faculty of Engineering and Informatics, Universitas. Panca Marga, Indonesia)
Ahmad Izzuddin (Electrical Engineering Department, Faculty of Engineering and Informatics, Universitas. Panca Marga, Indonesia)
Serlin Velinda (Electrical Engineering Department, Faculty of Engineering and Informatics, Universitas. Panca Marga, Indonesia)

Energy: Jurnal Ilmiah Ilmu-ilmu Teknik
Vol. 15 No. 2 (2025): ENERGY: JURNAL ILMIAH ILMU-ILMU TEKNIK (July-November 2025 Edition)

Maintaining skin health is essential, as poor skin conditions can lead to various diseases. To address this, early detection and classification of skin disorders is crucial. This study presents a deep learning-based Android application that enables users to detect and classify types of skin diseases through image input. The application integrates a Convolutional Neural Network (CNN) trained on labeled image datasets. The model achieved a training accuracy of 96% and validation accuracy of 83%. To provide a more comprehensive performance evaluation, metrics such as precision (87.75%), recall (84.29%), and F1-score (85.20%) were calculated. The evaluation was conducted using confusion matrix analysis based on eight skin disease classes. The implementation of CNN into an Android-based platform provides a practical and accessible tool for early skin disease detection and classification for the general public.