Animal Recognition From Natural Scene Images

With the development of imaging technology and research progress on wildlife monitoring, camera trapping becomes one of the best ways to record the presence and activity of mammals in a given area. The approach to monitoring wildlife can assist people in the community with decision making about preserving biodiversity. Camera trapping generates a huge volume of image data. In the past, experts analyzed such image data manually, which required domain knowledge and took significant time. In this project, we aim to develop an animal recognition system that can help analyze wildlife images, which record the presence of large mammals such as deer, moose, wolves, bear, etc., and automatically identify species of those animals. Such a system can help experts save significant time and better understand wildlife images and activities of animals around the certain area in a reasonable time. A Windows environment application will be built. This work will support ecologists to make a better decision on protecting and preserving the biodiversity in the province of Alberta.

Faculty Supervisor:

Osmar Zaiane

Student:

Partner:

Alberta Innovates - Technology Futures (Vegreville)

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

University of Alberta

Program: