Twenty years of AFORO: new developments and connections enhancing otolith research
Since 2003, with the aim of addressing different issues related to otoliths and the advancement of technologies, AFORO acronym of Anàlisi de FORmes d’Otòlits (otolith shape analysis), a computational environment with a set of tools including a website for this purpose, has gradually expanded its fun...
| Autores: | , , , , , , , , , |
|---|---|
| Formato: | artículo |
| Fecha de publicación: | 2025 |
| País: | España |
| Recursos: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/427452 |
| Acesso em linha: | https://hdl.handle.net/2117/427452 https://dx.doi.org/10.1016/j.fishres.2024.107242 |
| Access Level: | acceso abierto |
| Palavra-chave: | Otoliths Computer vision Neural networks Deep learning Artificial intelligence Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la imatge i del senyal vídeo Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial::Aprenentatge automàtic |
| Resumo: | Since 2003, with the aim of addressing different issues related to otoliths and the advancement of technologies, AFORO acronym of Anàlisi de FORmes d’Otòlits (otolith shape analysis), a computational environment with a set of tools including a website for this purpose, has gradually expanded its functionality. This implies different ways to measure otoliths in order to obtain different types of information of the same individual. The 2D otoliths shape description using different methods permits the comparison between classical methods and the Wavelet transform, used in the automatic classification system of AFORO. The 3D otoliths description in a few images opens the possibility of classifications using volumetric information, which they have performed well. The data of relationships between otolith and fish length in AFORO is used in predator-prey studies. Our website offers the possibility to display otolith georeferencing information interactively. Furthermore, the automatic classification system has been improved by incorporating geographic filters based on georeferenced data. The diverse information that AFORO offers on the otoliths opens the possibility to approach new studies or to improve the existing ones, combining this information or applying the most appropriate one to the specific problem. Classification cases with Wavelets that have been improved by introducing geographical area information will be presented. In addition, Deep Learning algorithms will be introduced by performing classifications with a small subset of otoliths of species which contain a sufficient number of specimens (>8). |
|---|