Imaging and processing of tree trunk

• Main Author: Hong, Yuan Ming
• Other Authors: Abdulkadir C. Yucel
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Electrical and electronic engineering::Antennas, wave guides, microwaves, radar, radio
• Online Access: https://hdl.handle.net/10356/139221
• Institution: Nanyang Technological University

Ground-Penetrating Radar (GPR) is traditionally used as a tool to detect objects beneath the geographical surface. It is a non-invasive method that uses electromagnetic waves (EM waves) in the microwave band as the propagation method. When the EM waves passes through a boundary of different permittivity level, part of the wave is reflected and received by the antenna. With the appropriate processing methods and algorithms, it is possible to derive the approximate distance an object is buried underneath the surface based on the reflection wave detected. GPR has since been explored further due to their non-invasive properties which can help monitor infrastructures and the environment. One such area being explored is the use of detecting and monitoring the health condition of trees. A tree health condition requires the examination of its roots, trunks, branches, and leaves. A circumferential B-Scan around the tree trunk is a common diagnostic method to determine the trunk’s health by imaging through EM waves or acoustic. In this report, planar B-Scans using EM waves were explored and seeks to analyse how does the various tree condition affects the planar B-Scan results. Common tree conditions such as cavities, cracks and decays are explored in this report. This method of exploration and analysis uses a simulation tool known as gprMax. gprMax is a software that can simulate EM waves with the appropriate modelling. The difference between 2D simulation and 3D simulation were analysed in this report too. A difference in peak reflection time between 2D simulation and 3D simulation was observed during B-Scan result analyses. Therefore, further investigations were conducted to better understand the peak time differences between 2D simulation and 3D simulation where the results can help to map the 2D simulations into 3D simulation using machine learning.