Tree species and aboveground biomass estimation using machine learning, hyperspectral and LiDAR data / Nik Ahmad Faris Nik Effendi
Above-ground biomass (AGB) and tree species classification using a combination of airborne hyperspectral and Light Detection and Ranging (LiDAR) can provide valuable and effective methods for forest management, such as planning and monitoring purposes. However, the identification process of tree spe...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/76575/1/76575.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Above-ground biomass (AGB) and tree species classification using a combination of airborne hyperspectral and Light Detection and Ranging (LiDAR) can provide valuable and effective methods for forest management, such as planning and monitoring purposes. However, the identification process of tree species and AGB estimation in tropical forest is quite challenging either by traditional or remote sensing methods due to the structure of forest type. Besides, only a few studies have applied using both combinations in the tropical forest. Therefore, the aim of this study is to classify and determine tree species and Above-ground biomass (AGB) and Carbon Stock using airborne hyperspectral and LiDAR at tropical forests. The objectives of this study to establish the AGB, Carbon Stock and species recognition using biophysical field data collection, to determine the individual tree species using different classifiers on hyperspectral data, and to estimate AGB using Random Forest (RF) and Artificial Neural Network (ANN). In this research, Object-Based Image Analysis (OBIA) method was applied on hyperspectral data to extract the crown of individual tree species for classification and estimation purposes. The result shows that Support Vector Machine (SVM) and Random Forest (RF) achieved the highest overall accuracy above 50% compared to other classifiers in the tropical forest. Besides, Artificial Neural Network (ANN) and Random Forest (RF) algorithm was used to predicted the AGB using different combination of variables. The best predicted selection using ANN is model 2 with produced RMSE = 24.117kg/tree and R2 = 0.999 which is two hidden layer. While the best predicted selection using RF is model 4 with mtry = p produced R2= 0.997 and RMSE=30.653kg/tree. Therefore, by using combination of field observation and remote sensing data with machine learning technique is reliable in forest management to estimate AGB in tropical forest. |
|---|