Advancing Fish Species Identification in Bangladesh: Deep Learning Approaches for Accurate Freshwater Fish Recognition

Deep learning, a subset of machine learning, has revolutionized scientific and environmental fields. In Bangladesh, the fishery industry faces sustainability challenges due to environmental changes, poor resource management, and limited knowledge of fish species. Fish fraud, including species substitution, further undermines the industry and jeopardizes consumer safety. Unlike other countries, Bangladesh lacks fish recognition technologies. To address this gap, this research proposes a deep learning system to identify Bangladeshi freshwater fish species. The system employs the VGGNet16 deep learning model and a custom dataset of 16,000 images of commonly consumed Bangladeshi freshwater fish. Through a mobile app, users capture fish images that are processed by the deep learning model to accurately identify the species. The app provides essential information such as generic name, local name, and scientific name. By improving fish identification, this research empowers consumers and promotes sustainable practices in the fisheries industry. The study fills the void of fish recognition technologies in Bangladesh by harnessing deep learning and a comprehensive dataset, enabling precise fish species identification, which benefits both consumers and fisheries management. The findings highlight the potential of machine learning in enhancing understanding and conservation efforts for Bangladesh's diverse freshwater fish species.