Ocean Plankton Biomass Estimation with a Digital Holographic Underwater Glider

Accurate quantitative plankton observation is significant for biogeochemistry and environmental monitoring. However, current equipment mostly shipborne, there a lack of long-term, large-scale, low-cost methods observation. This paper proposes solution to investigate using Seascan holographic camera equipped with “Petrel-II” underwater glider longer time sequence at larger scale. Aiming the new challenges low efficiency accuracy image processing after integrating imaging systems gliders, novel data analysis method applicable Digital Holographic Underwater Gliders (DHUG) proposed. The algorithm has following features: (1) high efficiency: breaks traditional hologram information extraction order, focusing only on key regions in minimizing redundant computation; (2) accuracy: applying Sobel variance focus plane significantly improves accuracy; (3) degree automation: by convolutional neural network, achieves fully automated observed data. A sea test South China Sea verified that proposed could greatly improve problems severe segmentation algorithms. It also proved DHUG survey great potential.