Development and Validation of a Tilapia Cutaneous Disease Detection System Using Machine Learning

Abstract

This study developed and validated a machine learning-based system integrated into a mobile application for detecting cutaneous diseases in tilapia. Focused on Oreochromis niloticus, the system utilized convolutional neural networks (CNN) trained with images from both local sources and online repositories. The application was tested with end users including farmers and IT experts, and evaluated using ISO 25010:2011 Software Quality Standards and the Technology Acceptance Model (TAM). Results showed high acceptability and accuracy, indicating that the system could serve as an effective early warning and disease management tool in aquaculture.