Examining the frontier of autonomous underwater vehicle image analysis for sea scallop incidental mortality
| Author(s) | Tipton, Hunter | |
| Date Accessioned | 2022-02-21T12:58:57Z | |
| Date Available | 2022-02-21T12:58:57Z | |
| Publication Date | 2021 | |
| SWORD Update | 2021-09-30T19:14:32Z | |
| Abstract | Historically, the Atlantic sea scallop (Placopecten magellanicus) fishery in the United States has undergone vast changes in management to combat stock volatility and fishing pressure. Today the American sea scallop stock has stabilized and is successfully managed via fishery independent surveys and stock assessment models. One of the factors affecting the stock assessment model for this fishery is incidental fishing mortality, or those individual scallops that experience mortality because of fishing effort but are not retained for use by the fishery. Recent studies sought to enumerate sea scallop incidental mortality on various seabed types. However, most of the scallops examined in these studies were inhabiting sandy bottoms more generally known as soft substrate. This study utilized 38,813 underwater images collected via Autonomous Underwater Vehicle (AUV) and a modified Multiple Before After Control Impact (MBACI) experimental design to examine incidental mortality on gravel and rock, or hard substrates within Closed Area I just north of the great south channel. Of the 31,972 scallops measured from these images 73% were annotated as lying on hard substrate and the highest measured value of incidental mortality was 6.70%. This result suggests that stock assessment values for incidental mortality remain conservative for hard substrate management areas. The results of this study also suggest that both sediment and sea scallop distributions may be more variable than previously understood within the spatial domain of a study site. | en_US |
| Advisor | Trembanis, Arthur C. | |
| Degree | M.S. | |
| Department | University of Delaware, School of Marine Science and Policy | |
| DOI | https://doi.org/10.58088/3krc-8q35 | |
| Unique Identifier | 1298584542 | |
| URL | https://udspace.udel.edu/handle/19716/30447 | |
| Language | en | |
| Publisher | University of Delaware | en_US |
| URI | https://login.udel.idm.oclc.org/login?url=https://www.proquest.com/dissertations-theses/examining-frontier-autonomous-underwater-vehicle/docview/2591387741/se-2?accountid=10457 | |
| Keywords | Atlantic sea scallop | |
| Keywords | Autonomous underwater vehicle | |
| Keywords | Deep learning | |
| Keywords | Image analysis | |
| Keywords | Multiple bore after control impact | |
| Title | Examining the frontier of autonomous underwater vehicle image analysis for sea scallop incidental mortality | en_US |
| Type | Thesis | en_US |