FishNet: Fish Classification using Convolutional Neural Network

Our main motive of this paper is to classify fish for people. Because many of us don't know the class of fish, name of fish. For that reason, we don't know the percentage of vitamins and nutritions of a fish properly and cannot understand which fish we should eat more. And this fish classification will also help the people who work in marine fishing and it is very important to know the classification of fish for them. Here we have identified six types of fish using the traditional algorithm along with Convolutional Neural Network (CNN) to classify fish. In traditional algorithms, we have implemented some remarkable and performable Machine Learning Algorithms. Convolutional Neural Network is considered as perfect and suitable for classification. That's why we use CNN to classify fish on our model. From various kinds of traditional algorithms, we got the best accuracy from Support Vector Machine(SVM) where the accuracy is 63.93% and our classification accuracy using Convolutional Neural Network is 88.96%.