Dried Fish Classification Using Deep Learning

Dried fish is a great procedure for fish reservation all over the world. Dried fish is evaluated as a choice food on the menu for a large number of Bangladeshi people. Dried fish is also considered as a proper origin of vitamins, minerals, and protein in the meal of people in numerous portions of the world along with Europe and Asia. 

Bangladesh is now the Asia’s fifth source of inland water fish, after only China and India (2020-2021). Dried fish is mostly produced from saltwater fishes captured by the fisherman and put up for sale all over the country by many steps of trading to arrive the customer. Thus, lots of fresher men and businessmen engaging in the trading business of dried fish. It is very crucial for the fresher man, businessman, and others people who want to involve this business to observe that; which type of fish drying will be profitable according to the market value and easier and low-cost drying method. The paper can assist fishermen, businesspeople, and common citizens in identifying the many types of dried fish. 

This set of data contains locally cognized dried fish like Bashpata, Chanda, Chapila, Chingri, Chouka, Dhela, Fesha, Ilish, Kachki, Loitta, Maya, Puti, Shundori, and Tengra. Some pictures of this dataset are collected by us then we have segmented and then augmented this dataset, this model is a trained Deep Learning and Convolutional Neural Network (CNN) based model for classifying dried fish. 

Keywords: Dried fish dataset, Augmentation, CNN, Image processing, Deep learning, Classification.