Fish Detection and Classification in IoT-Enabled Aquatic Environments with Deep Learning

Fish are an essential protein source, accounting for around 17% of the world’s animal protein consumption. In Malaysia, the fisheries industry plays a pivotal role in the economy, contributing approximately 11.22 billion Malaysian ringgit to the GDP in 2021. However, overfishing is a significant concern as it depletes marine life and disrupts ecosystems, potentially leading to a food shortage and unemployment. The Internet of Things (IoT) can transform fish farming, enhancing productivity, minimizing waste, and promoting sustainability. IoT devices supply data on various factors like fish behavior and water conditions, which can be utilized to develop AI models for early detection of inadequate conditions. The research validates the feasibility of smart fish farming based on IoT. Future enhancements include the ability to monitor fish growth and diseases, identify waterborne pathogens, and control pH and chlorine levels. It was shown that the results was at 97% for classifying the fish. The evolution of machine learning, deep learning, and transfer learning can facilitate the production of safe, high-quality, and abundant protein sources in well-regulated environments.