Detection of mature and immature oil palm from image Sentinel-2 using Google Earth Engine (GEE) / Nurul Ain Nabilah Sharuddin
Oil palm is crucial to ecology, the environment, and the economy. If improperly managed and monitored, unrestrained oil palm activities could contribute to deforestation, which can seriously affect the environment. Remote sensing provides a means to effectively detect and map oil palms from space. R...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/69181/1/69181.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Oil palm is crucial to ecology, the environment, and the economy. If improperly managed and monitored, unrestrained oil palm activities could contribute to deforestation, which can seriously affect the environment. Remote sensing provides a means to effectively detect and map oil palms from space. Recent developments in big data and cloud computing enable quick mapping on a wide scale. However, the use of cloud computing remains limited and challenging in Malaysia. Thus, this study used image Sentinel 2 processed in Google Earth Engine (GEE) to classify mature and immature oil palms' land cover in TDM Plantation, Terengganu. Four (4) machine learning algorithms are used in classification, such as Random Forest (RF), smile Classification and Regression Tree (smileCART), Gradient Tree Boost (GTB), and Minimum Distance (MD). Overall accuracy (OA) and kappa produced by Random RF, GTB, smileCART and MD were OA = 85.14%, kappa = 0.80, OA = 84.00%, kappa = 0.78 , OA = 83.42%, kappa = 0.77 and OA = 78.29%, kappa = 0.71 respectively, for 6 classes (water body, built up, mature, immature, bare land and forest). Therefore, image Sentinel-2 efficiently detected mature and immature oil palms' land cover using the RF algorithm implemented in GEE. |
|---|