AI routines for automated species classification and tracking by mobile crawler platform
2 pages, 2 figures
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | artículo de periódico biblioteca |
| Language: | English |
| Published: |
2023-01
|
| Online Access: | http://hdl.handle.net/10261/330880 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| id |
dig-icm-es-10261-330880 |
|---|---|
| record_format |
koha |
| spelling |
dig-icm-es-10261-3308802023-07-12T08:09:56Z AI routines for automated species classification and tracking by mobile crawler platform Monte, A. Marsiske, R. Ortenzi, L. Chatzievangelou, Damianos Costa, Corrado Thomsen, Laurenz Marini, Simone Aguzzi, Jacopo 2 pages, 2 figures Animal detection, classification and tracking as edge computing functionalities of mobile robotic platforms are increasingly relevant in marine ecosystem monitoring (Aguzzi et al., 2020; 2022). Time-series of geo-referenced counts for different species are crucial to train AI-based data processing algorithms. These will be executed on-board underwater robotic platforms, to deliver real-time, remote information on abundances and biodiversity. Crawlers are emerging mobile robotic platforms, either tethered to permanent infrastructures like cabled observatories, offshore industrial rigs, and mariculture assets (Danovaro et al., 2019), or moving autonomously along the seafloor for extended periods. Bearing cameras and complex sets of oceanographic and geochemical sensors, they can be used to consistently expand the radius of ecological monitoring of fixed cabled observatories by video-sweeping large seabed surfaces (Chatzievangelou et al., 2020), the benthic boundary layer (as the benthic-pelagic ecotone; Chatzievangelou et al., 2021), and the overlaying water column. Our objective is the automated, real-time image processing to classify and track multiple species, opening the pathway toward the creation of crawler on-board video-intelligence. Accordingly, manual classification of animals in videos acquired by the crawler Wally at the Barkley Canyon hydrates (900 m depth; NE Pacific) site of Ocean Networks Canada’s NEPTUNE observatory is ongoing, as a necessary step to create groundtruth datasets to train AI algorithms. Examples of key species (Figure 1) and AI tracking and classification (Figure 2) are provided Peer reviewed 2023-07-12T08:09:21Z 2023-07-12T08:09:21Z 2023-01 artículo de periódico Deep-Sea Life 20: 8-9 (2023) http://hdl.handle.net/10261/330880 en Sí none |
| institution |
ICM ES |
| collection |
DSpace |
| country |
España |
| countrycode |
ES |
| component |
Bibliográfico |
| access |
En linea |
| databasecode |
dig-icm-es |
| tag |
biblioteca |
| region |
Europa del Sur |
| libraryname |
Biblioteca del ICM España |
| language |
English |
| description |
2 pages, 2 figures |
| format |
artículo de periódico |
| author |
Monte, A. Marsiske, R. Ortenzi, L. Chatzievangelou, Damianos Costa, Corrado Thomsen, Laurenz Marini, Simone Aguzzi, Jacopo |
| spellingShingle |
Monte, A. Marsiske, R. Ortenzi, L. Chatzievangelou, Damianos Costa, Corrado Thomsen, Laurenz Marini, Simone Aguzzi, Jacopo AI routines for automated species classification and tracking by mobile crawler platform |
| author_facet |
Monte, A. Marsiske, R. Ortenzi, L. Chatzievangelou, Damianos Costa, Corrado Thomsen, Laurenz Marini, Simone Aguzzi, Jacopo |
| author_sort |
Monte, A. |
| title |
AI routines for automated species classification and tracking by mobile crawler platform |
| title_short |
AI routines for automated species classification and tracking by mobile crawler platform |
| title_full |
AI routines for automated species classification and tracking by mobile crawler platform |
| title_fullStr |
AI routines for automated species classification and tracking by mobile crawler platform |
| title_full_unstemmed |
AI routines for automated species classification and tracking by mobile crawler platform |
| title_sort |
ai routines for automated species classification and tracking by mobile crawler platform |
| publishDate |
2023-01 |
| url |
http://hdl.handle.net/10261/330880 |
| work_keys_str_mv |
AT montea airoutinesforautomatedspeciesclassificationandtrackingbymobilecrawlerplatform AT marsisker airoutinesforautomatedspeciesclassificationandtrackingbymobilecrawlerplatform AT ortenzil airoutinesforautomatedspeciesclassificationandtrackingbymobilecrawlerplatform AT chatzievangeloudamianos airoutinesforautomatedspeciesclassificationandtrackingbymobilecrawlerplatform AT costacorrado airoutinesforautomatedspeciesclassificationandtrackingbymobilecrawlerplatform AT thomsenlaurenz airoutinesforautomatedspeciesclassificationandtrackingbymobilecrawlerplatform AT marinisimone airoutinesforautomatedspeciesclassificationandtrackingbymobilecrawlerplatform AT aguzzijacopo airoutinesforautomatedspeciesclassificationandtrackingbymobilecrawlerplatform |
| _version_ |
1777667990320316416 |