Sentinel-2a Verileriyle Trabzon İli 2019-2020 Yılları Arasında Ortaya Çıkan Sınıflandırma Farklarının Çeşitli Algoritmalarla Değerlendirilmesi
Loading...
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Open Access Color
GOLD
Green Open Access
No
OpenAIRE Downloads
OpenAIRE Views
Publicly Funded
No
Abstract
Çevre durum raporları yayınlandıkları yıllar itibariyle ilgili şehrin ve çevresinin durumunu ortaya koyan önemli dokümanlardır. Trabzon İli ve çevresini kapsayan 2021 çevre durum raporundan yola çıkarak 1/100000 ölçekli çevre düzenleme planı (ÇDP) sınıflandırmasında yeşil alanların azalmakta olduğu fark edilmiştir. Bu çalışmada Trabzon İli yeşil alanların değişiminin 2019-2020 yılları arasında farklı kontrollükontrolsüz sınıflandırma yöntemleri ve makine öğrenmesi algoritmalarından rastgele orman (RO) algoritmasıyla incelenmesi sağlanmıştır. Değişimin izlenmesinde Normalize Edilmiş Fark Bitki Örtüsü İndeksi (NDVI), kontrollü, kontrolsüz ve RO algoritmasıyla sınıflandırma işlemleri gerçekleştirilmiştir. Çalışma alanının Sentinel-2A uydu veri setleriyle, 2019 yılı Mayıs ile 2020 yılı Mayıs ayları arasındaki, 10 m mekânsal çözünürlükteki bant kompozitleri kullanılarak ve NDVI işlemi gerçekleştirilerek sınıflandırma sonuçları değerlendirilmiştir. Sonuç olarak gerçekleştirilen kontrollü sınıflandırma işlemi sonucunda 2019 yılı için genel doğruluk oranı %100.0 iken 2020 yılı için genel doğruluk oranı %81.31 olarak elde edilmiştir. Uydu görüntülerinde makine öğrenmesi algoritmalarından RO algoritmasıyla sınıflandırma işlemi de gerçekleştirilmiştir ve genel doğruluk oranı %97.6 olarak elde edilmiştir.
Environmental status reports are important documents that reveal the situation of the relevant city and its surroundings as of the years they were published. Based on the 2021 environmental status report covering Trabzon Province and its surroundings, it has been noticed that green areas are decreasing in the 1/100000 scale environmental arrangement plan (EAP) classification. This study examined the change in the green regions in Trabzon Province with different controlled-uncontrolled classification methods and random forest (RF) algorithm, one of the machine learning algorithms, between 2019 and 2020. In monitoring the change, classification processes were carried out with Normalized Difference Vegetation Index (NDVI), controlled, uncontrolled, and RO algorithm. Classification results were evaluated using the Sentinel-2A satellite data sets of the study area between May 2019 and May 2020, using band composites with 10 m spatial resolution and performing NDVI processing. As a result, as a result of the controlled classification process, the general accuracy rate for 2019 was 100.0%, while the general accuracy rate for 2020 was 81.31%. Classification of satellite images was also carried out with the RO algorithm, one of the machine learning algorithms, and the overall accuracy rate was obtained as 97.6%.
Environmental status reports are important documents that reveal the situation of the relevant city and its surroundings as of the years they were published. Based on the 2021 environmental status report covering Trabzon Province and its surroundings, it has been noticed that green areas are decreasing in the 1/100000 scale environmental arrangement plan (EAP) classification. This study examined the change in the green regions in Trabzon Province with different controlled-uncontrolled classification methods and random forest (RF) algorithm, one of the machine learning algorithms, between 2019 and 2020. In monitoring the change, classification processes were carried out with Normalized Difference Vegetation Index (NDVI), controlled, uncontrolled, and RO algorithm. Classification results were evaluated using the Sentinel-2A satellite data sets of the study area between May 2019 and May 2020, using band composites with 10 m spatial resolution and performing NDVI processing. As a result, as a result of the controlled classification process, the general accuracy rate for 2019 was 100.0%, while the general accuracy rate for 2020 was 81.31%. Classification of satellite images was also carried out with the RO algorithm, one of the machine learning algorithms, and the overall accuracy rate was obtained as 97.6%.
Description
ORCID
Keywords
Sentinel-2a, NDVI, Sınıflandırma Yöntemleri, Classification Methods, Rastgele Orman, Random Forest
Fields of Science
0401 agriculture, forestry, and fisheries, 04 agricultural and veterinary sciences
Citation
WoS Q
Scopus Q

OpenCitations Citation Count
N/A
Volume
5
Issue
2
Start Page
78
End Page
88
PlumX Metrics
Citations
Scopus : 1
Captures
Mendeley Readers : 4
Google Scholar™


