| Autor: |
Tekić, Ivan, Jantol, Nela, Ljubić, Ivan, Neferanović, Andrea, Radun, Branimir, Tomljenović, Ivan, Žiža, Ivona, Kušan, Vladimir |
| Jazyk: |
angličtina |
| Rok vydání: |
2022 |
| Předmět: |
|
| Popis: |
Invasive plant species have long been recognized as a global threat to biodiversity, ecosystems and human health while their economic impact in Europe alone has been estimated at around 10 billion Euros per year. In Croatia, invasive plant species represent a severe problem for not only biodiversity but also cultural landscapes, while damages in agriculture and forestry exert more and more financial pressure because of management needs. This paper explores the potential of Sentinel imagery as a tool for understanding, tracking and quantifying invasive plant species in Croatia and represents the first step in the development of a reliable, near-real-time service for their tracking. Detection of selected plant species was carried out by identification of spectral discrepancies between multidate images which will be calculated through differencing the same spectral bands or indices for case studies where ground truth data were collected. Through the application of machine learning algorithms, computed results were analysed with the aim of obtaining the most significant variables that can differentiate invasive plant species from other types of vegetation. The research fills the gap in the application of these techniques in Croatia and should open a path for an accurate way to determine spatial cover, extent, and trend of invasive processes. It also provides support and development of optimal mitigation and eradication approaches with possible application in other EU Member States. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
