Projects in image analysis and computer graphics
Ela Claridge
2013/2014 Final Year Projects
2014 MSc projects
I am most keen to supervise projects in image processing and image analysis.
I can also supervise computer graphics projects.
If you have ideas of your own in these areas, I shall be very happy to discuss
them with you.
Some of the projects below are related to current research in the School. Many
involve mathematics and the students choosing them will have to be prepared to
master the necessary background and techniques - if needed I shall provide
support and guidance.
Image processing and image analysis
Computer graphics and visualisation
Medical image analysis
Medical image analysis tries to find descriptions of important diagnostic
signs in medical images and to use them for diagnosis. Computer-derived
descriptions are often more accurate and repeatable than those derived
by human observers.
The descriptions can include shape, colour, pattern, texture, and many
other image features.
Several projects are available, including:
- Analysis of skin moles, to extract descriptions which help in diagnosis
of melanoma, a skin cancer.
- Detection of important structures in images of the eye
Tracking the movement of blood cells from microscopy video recordings
Using high quality video camera and a microscope, bioscientists record the movement of the blood cells in the vein. Slowly moving cells may stick to the vein surface and eventually form a clot, which could be a health risk. It is therefore important to be able to track the individual cells, assess their speed, and establish whether there are any slow moving cells.
This project will involve image processing and analysis to detect all the cells in each video frame, and then to track each cell from frame to frame. From the change of the position in the frame and the speed of the recording the speed of the movement can then be computed.
Jigsaw puzzle
This project will develop a program to solve simple pictorial jigsaw puzzles.
A digital image will be subdivided into simple square blocks. By using
a mixture of pattern matching techniques applied to pictorial patterns,
and search algorithms, the puzzle will be put it together. This is quite
a challenge!
This project is particularly suitable for CSAI students.
Iris detection for biometric analysis
Iris recognition is one of the most robust biometric methods. Before recognition
it is necesary to detect and segment the key structures in images of the iris:
the iris outline and the pupil. This is the key aspect of the project.
Students who like mathematics will be able to attempt iris recognition. Iris
images are available from the CASIA iris database.
Creating realistic animations of
dynamic natural phenomena (cloud, fire, smoke, etc.)
A computer graphic technique called "particle cloud animation" can be
used to create realistically looking scenes showing "amorphous" things
such as fire or clouds. It's two key components include (1) probabilistic
generation of particle paths; and (2) model-based changes in particle
colour. On the top of that, clever implementation is required, to
simultaneously move in 3D a huge number of particles in real time.
Computer graphics rendering of
seashells
Seashells can be very beautiful, but they are also highly structured objects.
Because of this, they can be relatively easily generated using computer
graphics.
This project will involve the development of the "shell growing" program
which would generate
various shell types. These shells can then be viewed and manipulated.
Maintained by Ela Claridge
Last update: 13 September 2013