Bioluminescence refers to the production of ecologically functional light by living organisms. It is widespread in the marine environment, where it occurs in a broad range of phyla. The deep pelagic ocean is the largest biome on earth and is chronically under-sampled. Underwater camera systems offer a rapid sampling method for this zone. In the current study, the ICDeep (Image Intensied Charge Coupled Device for Deep-sea research) profiler was used to record digital video of bioluminescent organisms through the deep water column. This work proposes a new automatic detection and counting methods for bioluminescent organisms represented as ashes within video data acquired through the camera based on computer vision algorithms. The proposed method has been validated by a ground truthed sequence of bioluminescent events resulting in a high correct detection rate and real time processing execution.
|Title of host publication||Proceedings|
|Subtitle of host publication||2014 ICPR Workshop on Computer Vision for Analysis of Underwater Imagery (CVAUI)|
|Number of pages||8|
|Publication status||Published - 20 Nov 2014|
|Event||2014 ICPR Workshop on Computer Vision for Analysis of Underwater Imagery - Stockholm, Sweden|
Duration: 24 Aug 2014 → 24 Aug 2014
|Conference||2014 ICPR Workshop on Computer Vision for Analysis of Underwater Imagery|
|Abbreviated title||CVAUI 2014|
|Period||24/08/14 → 24/08/14|
- bioluminescent organisms
- deep sea
- under- water computer vision
- underwater object recognition.
- video tracking
Mazzei, L., Marini, S., Craig, J., Aguzzi, J., Fanelli, E., & Priede, I. G. (2014). Automated video imaging system for counting deep-sea bioluminescence organisms events. In Proceedings: 2014 ICPR Workshop on Computer Vision for Analysis of Underwater Imagery (CVAUI) (pp. 57-64). IEEE Explore. https://doi.org/10.1109/CVAUI.2014.15