Fish Fry Dataset (FFD): used for stocking density control and health assessment

This study presents a dataset of 1101 images of largemouth bass (Micropterus salmoides) fry, specifically designed for small target detection under dense scenes. Each image contains a variable number of fish fries ranging from 20 to 80 individuals. To facilitate health assessment in the aquaculture context, a small number of dead fries are also included in each image. The entire dataset is annotated with a total of 51119 live fish fry and 3586 dead ones. In addition, among the 80 images depicting high density scenarios, there is a prevalence of complex situations such as overlapping, occlusion, and adhesion, which brings challenges to the small target detection task. The dataset is annotated using the labelimg tool and converted to the COCO format. The dataset can be applied to a variety of scenarios including seedling rearing, fry purchase and sale, and survival assessment. It is also valuable for biomass estimation and aquaculture density control. In summary, this dataset provides an invaluable resource for the research community, advancing the study of fry counting and fish population health, thereby contributing to the development of intelligent aquaculture.