Go to  Advanced Search

Analysis of historic aerial photographs for ecological management using object-based approaches

Show full item record

Files in this item

Files Size Format Description   View
ubc_2009_fall_morgan_jessica.pdf 1.904Mb Adobe Portable Document Format   View/Open
 
Title: Analysis of historic aerial photographs for ecological management using object-based approaches
Author: Morgan, Jessica Lindsay
Degree Master of Science - MSc
Program Forestry
Copyright Date: 2009
Publicly Available in cIRcle 2009-10-01
Abstract: Aerial photographs are a crucial tool for ecological monitoring and management. New approaches for aerial photograph analysis are needed because of several existing (and anticipated) challenges associated with traditional analysis techniques. The goal of this thesis is to provide a synthesis of the valuable, and often unique, ecological information available from aerial photographs, and to explore the utility of novel image analysis approaches to extract this information. This research is organized to reflect the future of aerial photography as a discipline. In chapter two, I review the benefits and challenges of using aerial photographs for ecological management. The traditional framework used to classify aerial photographs, as well as sources of error are described within the context of the diverse ecological questions that can be addressed using aerial photography. The need for new approaches to analyze aerial photographs is emphasized throughout this chapter. In chapter three, I compare manual interpretation to an automated approach (combining object-based analysis and classification tree modeling), for five classification schemes routinely used in British Columbia. Automated approaches hold potential for replicating certain aspects of the manual process (such as the delineation of polygons), as automated and manually-delineated objects display few statistical differences. Automated classification accuracy is highly variable, with individual class accuracies ranging from 0–74%; however the overall accuracy of several classification schemes exceeded 60%, suggesting certain schemes are well suited to automated analysis. In chapter four, object-based analysis is applied to historic aerial photographs to better quantify spatial heterogeneity, a concept fundamental to the field of landscape ecology. My results suggest sixteen independent factors are needed to describe baseline levels of landscape heterogeneity, including several factors not previously identified by the discipline of landscape ecology. Lastly, in Chapter five, the significance of this thesis for resource management, remote sensing, and landscape ecology is highlighted. Further avenues of research are also discussed.
URI: http://hdl.handle.net/2429/13462

This item appears in the following Collection(s)

Show full item record

All items in cIRcle are protected by copyright, with all rights reserved.

UBC Library
1961 East Mall
Vancouver, B.C.
Canada V6T 1Z1
Tel: 604-822-6375
Fax: 604-822-3893