Tracking and mapping underwater bioluminescent displays using snakes

The search for recognizable and explicable patterns in the organization of pelagic marine ecosystems is critical to developing the predictive capability of characterizing natural, compared to anthropogenic, perturbations in the marine ecosystem. Bioluminescence mapping is an innovative solution to this very challenging problem of how to characterize the in situ spatial relationships of organisms in the three-dimensional fluid environment of the open ocean. Based on the unique spatial and temporal patterns associated with stimulated bioluminescent displays, organisms as small as 50 micrometer dinoflagellates can be identified and mapped in three dimensional space. This paper presents a novel, automated approach to tracking and mapping recorded underwater bioluminescent organisms based on their display patterns. In this approach, a computational technique, known as snakes, assists in image segmentation, labeling, tracking and mapping. The results can be used to help researchers identify luminous events and organism species, and extract features such as duration, size, and coordinates of the point of impact from the collected video data. Future research will focus on using this information to create a more complex multi-dimensional model of each display and to automatically identify each organism