Naslov (srp)

Interaktivna identifikacija evropskih vrsta egzofitofagnih gusenica kroz upravljanje deskriptivnim podacima

Autor

Trajković, Aleksandra, 1996-

Doprinosi

Žikić, Vladimir, 1969-
Tomanović, Željko, 1966-
Ilić-Milošević, Marijana, 1983-
Popović, Miloš, 1987-
Stanković, Saša, 1982-

Opis (srp)

The growing use of computer vision and deep learning for species identification has significantly sped up the collection of biodiversity data worldwide, including data on the diverse morphological and ecological group of exophytophagous caterpillars. While mobile apps are increasingly replacing extensive conceptual literature and are widely used in both scientific and professional circles, as well as by the public, they often miss a crucial aspect of fully understanding a taxon: the biology and ecology of a species as an entity functioning within a complex system of biotic and abiotic factors. This dissertation explores a somewhat opposing question: do new technologies compromise, or do they enhance, traditional species treatments at the trait level?In this work, species identification is approached as a multidimensional concept that relies not only on biological traits but also on how these traits are interpreted through cognitive processes. Three trait data matrices were developed, containing standardized morphological, ecological, and behavioral descriptors for selected species. Each species set, and each matrix, serves to demonstrate, accurately quantify, and prepare data for statistical analysis—whether addressing polymorphisms and polyphenisms within the Lepidoptera order, the relationship between habitat and habitus, or the use of expert systems in species identification.The goal of the dissertation is to show that visual signals and contextual information used in the identification process can be further refined through a combination of manual and semi-automatic approaches to gathering and processing descriptive data, the use of crisp and fuzzy coding for descriptors, managing variability at multiple levels, and selecting appropriate categorical counterparts for predictive models. Ultimately, the standardized and refined data were used to create an interactive web application that, through a filtering system, enables species identification and links each taxon to its role within the ecosystem.

Opis (srp)

Biografija autora: list 135.Bibliografija: listovi 108-130. Datum odbrane: 12.04.2025. Entomology

Jezik

srpski

Datum

2024

Licenca

Ovo delo je licencirano pod uslovima licence
Creative Commons CC BY-NC-ND 3.0 AT - Creative Commons Autorstvo - Nekomercijalno - Bez prerada 3.0 Austria License.

http://creativecommons.org/licenses/by-nc-nd/3.0/at/legalcode