Abstract- Recommending a way for the automated detection of fishspecies using computer vision is the target of this paper. High accuracy offish classification is widely used in order to understand fish behavior for thefishermen and the fishing community by large. Endangered species in waterbodies concern multiple institutions. The existing methods addressclassification of fish only as a classification model which is not real timebecause implementing real time classification through an usable applicationposes challenges. This method uses techniques based on Convolutional NeuralNetworks, Deep Learning and Image Processing to achieve maximum accuracy and itensures a considerably user friendly system with accuracy improvements.
Keywords- Fish Species Detection,Computer Vision, Deep Learning, Convolutional Neural Network