The Freshness of the Fish Eyes Dataset

Published: 31 January 2022| Version 1 | DOI: 10.17632/xzyx7pbr3w.1
Eko Prasetyo,


The Freshness of the Fish Eyes (FFE) dataset is a dataset for classifying freshness of fish based on eye images. This dataset consists of 4392 images of fish eyes, consisting of eight fish species; each species consists of highly fresh (day 1 and 2), fresh (day 3 and 4), and not fresh (day 5 and 6). The eight fish species as follows Chanos Chanos (500 images), Johnius Trachycephalus (240 images), Nibea Albiflora (421 images), Rastrelliger Faughni (769 images), Upeneus Moluccensis (792 images), Eleutheronema Tetradactylum (240 images), Oreochromis Mossambicus (625 images), and Oreochromis Niloticus (805 images). Each species is divided into three levels of freshness so that all 24 classes. The number of images for each class varies as follows: Chanos Chanos (168, 162, 170 images), Johnius Trachycephalus (80, 80, 80 images), Nibea Albiflora (173, 125, 123 images), Rastrelliger Faughni (336, 216, 217 images ), Upeneus Moluccensis (310, 252, 230 images), Eleutheronema Tetradactylum (80, 80, 80 images), Oreochromis Mossambicus (289, 174, 162 images), and Oreochromis Niloticus (328, 231, 246 images). For use this dataset, please cite this related article: Prasetyo, E., Purbaningtyas, R., Adityo, R.D., Suciati, N., & Fatichah, C. (2022). Combining MobileNetV1 and Depthwise Separable Convolution Bottleneck with Expansion for Classifying the Freshness of Fish Eyes. Information Processing in Agriculture.



Institut Teknologi Sepuluh Nopember, Universitas Bhayangkara Surabaya


Computer Science, Aquaculture