Modelling the representation of tree root properties within ground penetrating radar (GPR) imagery

Authors: Justin Miron*, Urban Forest Research and Ecological Disturbance (UFRED) Group, Ryerson University, Andrew A Millward, Urban Forest Research and Ecological Disturbance (UFRED) Group, Ryerson University
Topics: Biogeography, Urban Geography, Environment
Keywords: urban forest, tree roots, detection, GPR, tree protection
Session Type: Paper
Day: 4/4/2019
Start / End Time: 3:05 PM / 4:45 PM
Room: Marshall South, Marriott, Mezzanine Level
Presentation File: No File Uploaded

Effective urban forest management, and especially tree protection, requires accurate knowledge of the locations of underground tree roots. Ground Penetrating Radar (GPR) is a remote sensing method that can enable accurate detection of tree roots buried under the ground. GPR produces a two-dimensional, vertically-oriented, image that represents the differential energy reflections of various sub-surface objects, including tree roots. These images can be very complex and, therefore, it can be difficult to accurately detect root locations within them, either manually or automatically. This project develops a predictive model that links several material properties of roots, such as moisture-content, overall mass, and orientation relative to the GPR scan, to the signal properties of their associated energy reflections within a GPR scan. The developed model is then used as an input to an automated root detection software program. Such a model will enhance the accuracy and precision of root location detection for GPR scans obtained in a variety of soil conditions in the field. The refinement of methods of automatic root detection in GPR imagery contributes to the effectiveness of urban forestry practice by delivering a non-invasive and cost-effective approach to the identification of root locations.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login