The fisheries sector plays a vital role in the economy and food security of Bangladesh. Bangladesh is one of the leading countries in inland fish production. Bangladesh gains sustainable economic benefits from aquaculture and fisheries. This sector made a significant contribution to the GDP and ensures employment for approximately 18 million people. Fish is one of the primary sources of protein for the population, accounting for more than 60% of the country’s animal protein intake. Efficient fish species identification is relevant to sustainable fisheries management, smart aquaculture, and food authenticity. This dataset includes 2455 clear images of seven frequently consumed freshwater fish in Bangladesh: Shrimp, Prawn, Mola Carplet, Dwarf Gourami, Swamp Barb, Stinging Catfish, and Mystus Catfish. All data were collected from the fish-rich areas in Bangladesh—Netrokona and Bogura. Data samples were collected from ponds, rivers, and fish markets, both natural and commercial sources. The diverse environment provides variation in lighting, background, and orientations, which highlights the real-world complexity for image classification. Each species is classified as scientific, local, and English names for accurate recognition. The dataset is suitable for research in smart aquaculture, including fish identification and species recognition. The collected data allows building machine learning models for image classification and allows fine-tuning previous models for local applications. The dataset includes real-world variability, which may support the generalization and robustness of machine learning models. This dataset provides a strong foundational resource for academic research and practical implementation in smart aquaculture. This dataset aims to contribute to sustainable fisheries and similar ecosystem development.
Biswas et al. (Wed,) studied this question.