An intelligent vision system for wildlife monitoring

Fagg, Ashton (2012) An intelligent vision system for wildlife monitoring. [USQ Project]

[img] PDF
Fagg_2012.pdf

Download (3196Kb)

Abstract

The understanding of animal behaviour in response to human development is vital for sustainable management of ecosystems. Existing methods of monitoring wildlife
activity fall short in facets pertaining to accuracy, accessibility, cost and practicality.

To address this level of crudity, recent technological advances have led to the development of electronic, autonomous wildlife monitoring solutions. Whilst these developments improve the overall experience in some areas, there is still much to be desired. This dissertation aims to outline the development of an accessible, affordable, intelligent vision-based technique which addresses limitations of existing monitoring methods.

A signal processing methodology was investigated, developed and implemented. The development of this methodology included the investigation of two distinct facets of
computer vision - image segmentation and event classification.

The existing literature was explored, and several image segmentation techniques were explored. Upon further investigation, the Gaussian Mixtures Model was selected in two forms - per pixel modelling (Zivkovic 2004) and a compressive sensing based method(Shen, Hu, Yang, Wei & Chou 2012). Each method was evaluated in terms of real time
capabilities and accuracy to provide basis for recommendation of the method presented in the prototype.

Upon evaluation, it was discovered that the proposed compressive sensing based method was a suitable prototype and recommendations regarding the implementation and com-
missioning of the system were made. Furthermore, possible avenues for further research and development were explored.


Statistics for USQ ePrint 23111
Statistics for this ePrint Item
Item Type: USQ Project
Refereed: No
Item Status: Live Archive
Faculty / Department / School: Historic - Faculty of Engineering and Surveying - Department of Electrical, Electronic and Computer Engineering
Supervisors: Leis, John; Kottege, Navinda
Date Deposited: 27 Feb 2013 00:25
Last Modified: 27 Feb 2013 00:25
Uncontrolled Keywords: wildlife monitoring; vision system; signal processing
Fields of Research : 09 Engineering > 0906 Electrical and Electronic Engineering > 090609 Signal Processing
URI: http://eprints.usq.edu.au/id/eprint/23111

Actions (login required)

View Item Archive Repository Staff Only