A carregar...
Publicação
Monitoring and evaluation of the best mapping approach of the environmental impacts and land cover changes caused with the reveal of the quarry using the landsat and UAV images
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.61 MB | Adobe PDF |
Autores
Resumo(s)
The major incidents caused from mine waste and poor monitoring of the quarry sites in Serbia has caused the formation of a large number of contaminated areas that are causing environmental and social impacts and land cover changes. Related impacts in the European countries has initiated the formation of the environmental legislation for the prevention of the quarry risks on the environment. This project evaluates the EU MWD Protocol on the real case of the quarry site “Gradinje”. The protocol is linked with the GIS system and showed the presence of the contamination, which later was mapped using the most adequate classification approach. Pixel based image analysis approaches for classifying land cover classes are compared using three supervised classification algorithms: maximum likelihood (ML), support vector machine (SVM) and random forest (RF). Concerning the comparison between different classification predictions, the McNemar test indicated that the observed difference between the two machine learning algorithms (RF and SVM) was not statistically significant (p = 0.2278 > 0.05). However, there was not a statistical significant difference (p > 0.05) in classification accuracy between map produced using SVM algorithm and RF algorithm. Classification based on SVM and RF algorithms produced map that is more visually adequate depiction of land cover classes than that produced by ML algorithm. Overall, two machine learning algorithms seems superior than ML algorithm in land cover classification using medium spatial resolution imagery in term of performance based on classification accuracy.
Descrição
Dissertation submitted in partial fulfilment of the requirements for the degree of Master of Science in Geospatial Technologies
Palavras-chave
Environmental risk Geographical Information Systems Landsat Imagery Maximum Likelihood Quarry risk assessment Bespilotna letjelica Daljinska detekcija Ekološki rizik Evropska direktiva o rudarskom otpadu