Tree height estimation using field measurement and Low-Cost unmanned Aerial Vehicle (UAV) at Phnom Kulen national Park of Cambodia

Name: Ly Mot

LCSH: Burapha University.Faculty of Geoinformatics

Classification :.DDC: 530.8

LCSH: Measurement

LCSH: Physical measurements

LCSH: Drone aircraft

Abstract: Tree height estimation is one of the most important parameters used to quantify timber resources. Among others it is used to evaluate the ecological and economic value of forest stand, to calculate the individual and number of stand volumes, and to estimate the forest inventory. In order to update information about forests. This helps local, regional, or national authorities to take decisions and manage the forest. Most tree estimations with Light Detection and Ranging (LiDAR) have been used successfully over the recent decades. In contrast to LiDAR, estimation of tree height derived Canopy height model (CHM) has been applied with low-cost UAV with acceptable accuracy which is used onboard GPS to obtain a high accuracy of CHM. This research aims to estimate, and evaluate tree height from high resolution images of low-cost UAVs. The influence of different flight attributes, point cloud densities, extraction methods, photogrammetry products, and point cloud classification are discussed. The estimation of tree height was performed by two extraction methods, photogrammetry product, and point cloud classification. Each was divided into five groups: CHM from the point could classifications, CHM from photogrammetry products-customized with georeferenced methods, CHM from photogrammetry products-defaults with georeferenced methods, CHM from photogrammetry products-defaults without georeferenced methods, and CHM from photogrammetry products- customized without georeferenced methods. Tree heights were obtained from the field with buffering distances of 0 cm, 50 cm, 100 cm, 150 cm, and 200 cm. In total 50 measurements were taken and analyzed in the present study First, the results of tree height extraction were successfully taken from UAV data and point cloud classification. In contrast, photogrammetry products produced tree height estimation with extreme bias. In addition, high point cloud densities from 50 m flights provided good data to remove point cloud outliers. The highest R2 was 0.60. During 200 m flights, R2 of 0.50 was the highest. Additionally, sample paired t-test, tree height estimation, and ground data from 50 m flight were not statistically significant different. In sum, this proposed method is possible for open terrains less than 12 m. It is limited by the design of the pipe meters as the measurement of height and cashew leaves was challenging. Regarding the performances of tree height estimation from UAV and field measured, we proved that the workflow of UAV is faster and more effective than field measured which required less times and resources

Burapha University. Library

Address: CHONBURI

Email: buulibrary@buu.ac.th

Name: Hong Shu

Role: Principal advisor

Name: Kitsanai Charoenjit

Role: Co-advisor

Name: Haoran Zhang

Role: Co-advisor

Created: 2021

Modified: 2022-06-02

Issued: 2022-06-02

วิทยานิพนธ์/Thesis

application/pdf

CallNumber: Th 530.8 Ly981T

eng

DegreeName: Master of Science

Level: Master's Degree

Descipline: Geoinformatics

Grantor: Burapha University

©copyrights Burapha University

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	62910209.pdf	4.17 MB	3	2024-10-16 19:40:57

ใช้เวลา

0.045917 วินาที