Land use land cover classification from different classifiers in Kedah between 2014 and 2021 using Google Earth engine / Noorain Ahmad
Over time, a growing human population that accelerates land use and land cover (LULC) change has placed a massive burden on natural resources. Changes in LULC have become an essential issue for decision makers and environmentalists. Monitoring and evaluating LULC changes over large areas become crit...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/69367/1/69367.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Over time, a growing human population that accelerates land use and land cover (LULC) change has placed a massive burden on natural resources. Changes in LULC have become an essential issue for decision makers and environmentalists. Monitoring and evaluating LULC changes over large areas become critical. Understanding the functional diversity of machine learning classifiers is important due to increased geospatial data from satellite remote sensing. The potential of Google Earth Engine (GEE) as cloud-based computing to know the changes of the map area in a long period is interesting to study. Therefore, this study aims to evaluate the LULC classification map using different classifiers in Kedah between 2014 and 2021 conducted on Google Earth Engine. The objective of this study is i) to classify LULC carried out on the Google Earth Engine Platform using three (3) different classifiers (Random Forest, smile CART, and Minimum Distance) for Landsat 8 images in Kedah between 2014 and 2021, ii) To compare their performance for three (3) classifiers using accuracy assessment, and iii) to produce land use land cover maps for the years 2014 and 2021 for each classification. Landsat 8 images are obtained from GEE, and all the processing involved is done on this platform. The study prove that the best classifier was Random Forest and the OA = 80.50%, kappa = 0.73 for 2014 while in year 2021 OA = 80.88%, kappa 0.75. This study shows the GEE cloud platform's efficiency in generating spatial temporal classification maps with high accuracy and takes a short time, and is easy to modify. The final output for this study was the LULC maps for the years 2014 and 2021 will benefit local authorities and policy makers for their planning and sustainable management. |
|---|