Image-Based Monitoring of Jellyfish Using Deep Learning Architecture

Jellyfish blooms have caused great damage to the fishery industry. In efforts to solve this problem, various systems to remove jellyfish have been proposed. This letter presents preliminary results of applying an image-based jellyfish distribution recognition algorithm to increase the efficiency of an existing jellyfish removal system. By using a convolutional neural network and dedicated image processing techniques, the experimental results show reasonable performance for real-world application.