A vision-based detection algorithm for moving jellyfish in underwater environment

This paper proposes a novel vision-based jellyfish detection algorithm in underwater environment. Since jellyfish, especially Aurelia aurita, is translucent and amorphous when projected to the 2-dimensional image, its boundary is indistinct and detecting jellyfish is very challenging. To detect moving jellyfish, two types of features that present the translucent appearance and swimming motion of jellyfish, respectively, are proposed. The proposed algorithm tracks and recognizes swimming jellyfish. The feasibility test of the proposed algorithm shows that it can detect jellyfish in underwater environment.