|
2007 Projects
2006 Projects
2005 Projects
2004 Projects
2003 Projects
Dr Du Huynh has more than 15 years of research experience in computer
vision.
Her research areas include shape from motion, 3D shape reconstruction,
visual tracking, video and image analysis.
A few projects in computer vision are offered in 2008. Some of
of them may be jointly supervised with staff members of the
School. If you have in mind a computer vision research topic
that is not listed below, I would be interested to hear from
you. Please note that I will be able to supervise up to 4
projects in any one year.
With appropriate adjustment, any
of the projects below
could be suitable for a BE(SE) final year project (12 points), an
Honours Research Project (24 points), or a MSc project (24 points).
Experience has shown that it can
be very beneficial for research students
to have a group of people with related interests to share ideas with. A
student undertaking any of the projects below is expected to join the
Computer Vision Research Group and will be expected to attend and
contribute to group meetings and discussions. Such a student will be
housed in the Computer Vision Research Group Laboratory in Room 2.09 of
the Computer Science building.
You are also strongly advised to take the CITS4240 Computer Vision
unit offered in the first semester.
- Image segmentation
Image segmentation is the process of partitioning the image into
meaningful regions. Each region is simply a group of connected pixels
with similar properties, such as grey levels, colours, textures, etc.
The importance of image segmentation should not be over-looked, as
many computer vision applications, such as pattern recognition, scene
understanding, and object counting, involve image segmentation as one
of their subtasks. In this project, you will study and implement a
recent image segmentation technique proposed by Nock and Nielsen [2].
A recently thorough literature review should also be included in your
thesis to demonstrate your understanding of several other image segmentation
techniques that have been proposed in the past, e.g., region growing
[1], the watersheds algorithm [3], etc. One of these methods should
be implemented and compared with the technique of Nock and Nielsen on
a variety of images.
This project will be suitable to students who have some basic
knowledge in statistics.
References:
[1] Rafael C. Gonzalez and Woods: "Digital Image Processing", 2nd Edition,
Prentice Hall, 2002.
[2] Richard Nock and Frank Nielsen: "Statistical Region Merging", IEEE
Transactions on Pattern Analysis and Machine Intelligence, vol. 26,
no. 11, Nov. 2004.
[3] Luc Vicent and Pierre Soille: "Watersheds in Digital Spaces: An
Efficient Algorithm Based on Immersion Simulations", IEEE Transactions
on Pattern Analysis and Machine Intelligence, vol. 13, no. 6,
Jun. 1991.
- Tracking Multiple Humans in Video Sequences
The tracking of multiple humans in video sequences is important in
many computer vision tasks, such as video surveillance, event
inference, activity classification and analysis. A successful
tracking technique will require the modelling of background that can
adapt to lighting changes and the modelling of human shape. This
project will have a lot of fun to those students interested in
capturing and processing video data. Detailed literature review on
related techniques in background and human shape modelling must be
included in your thesis. The tracking will be assumed to be done
off-line so implementation in either C or Matlab will both be
acceptable. If you are interested in this project,
please see [1] and follow the references in that paper.
References:
[1] Tao Zhao and Ram Nevatia: "Tracking Multiple Humans in Complex
Situations", IEEE Transactions on Pattern Analysis and Machine
Intelligence, vol. 26, no. 9, Sep. 2004.
- Outlier and Motion Detection using PCA
Principal Component Analysis (PCA) has been widely used for
dimension reduction in many problems. In the computer vision area, it
has also been used for the representation of shape, motion, and
appearance. PCA works on analysing the covariance matrix of the data
to extract the principal components. This can be done either via the
singular value decomposition (SVD) or eigen-decomposition of the
covariance matrix. This process is rather expensive and so an
iterative approach is often more desirable. A drawback of PCA is that
it cannot handle the presence of outliers in the data. In this
project, a robust PCA method proposed by Torre and Black [1] will be
studied and implemented for the detection of outliers in data and for
motion segmentation from video images. Those students aiming to get a
first class Honours should also accomplish a thorough literature
review on other outlier detection techniques (e.g., RANSAC,
LMedS).
Basic knowledge in linear and matrix algebra will be
essential.
References:
[1] Fernando de la Torre and Michael J. Black:
"Robust Principal Component Analysis for Computer Vision",
International Conference on Computer Vision, Vancouver, Canada, July
2001.
 Return
to the list of 4th year projects
|
