Course analysis 2D1257 Visualisation, 4p Spring 2003

Course data

Staff

Course responsible
Kai-Mikael Jää-Aro
Other teachers:
Sarah Amandusson
Ulf Andersson
Sten-Olof Hellström
Nils Smeds

Extents

16 h lecture, 10 h supervised lab

A laboratory exercise worth 2 points, one written exam worth 2 points, four exercises giving bonus points on the exam.

25 students, (4 D, 1 Dr, 1 FK, 1 M, 18 IM)

Course literature

OpenDX - Paths to Visualization by David Thompson, Jeff Braun and Ray Ford.
"Visualization of Scattered Meteorological Data: Study of Severe Rainfall Events in Northwestern Peru", by Lloyd A Treinish, Proceedings of the 1996 IBM Visualization Data Explorer Symposium, 1996.
"Why Should Engineers and Scientists Be Worried About Color?", by Bernice E Rogowitz and Lloyd A Treinish, 2001.
"Perceptual Techniques for Scientific Visualization", by Christopher G Healey, SIGGRAPH '99 Course Notes #6, 1999.
"Imaging Vector Fields Using Line Integral Convolution", by Brian Cabral and Leith Leedom, Computer Graphics 27(4), 1993.
"Visualizing Multivalued Data from 2D Incompressible Flows Using Concepts from Painting", by R M Kirby, H Marmanis and D H Laidlaw, IEEE Visualization, 1999.
"Marching Cubes: A High Resolution 3D Surface Construction Algorithm", by William E Lorensen and Harvey E Cline, Computer Graphics 21(4), 1987.
"Conveying the 3D Shape of Smoothly Curving Transparent Surfaces via Texture", by Victoria Interrante, Henry Fuchs and Stephen M. Pizer, IEEE Transactions on Visualization and Computer Graphics, 3(2), 1997.
"Showing Shape with Texture—Two Directions are Better than One", by Sunghee Kim, Haleh Hagh-Shenas and Victoria Interrante, 2003.

Student performance

After the make-up exam in April, the results were as follows:
U
   
3
(G)
4
   
5
(VG)
D 1 0 1 2
Dr 0 0 0 1
FK 0 0 - 1
IM 0 6 5 4
M 0 0 0 0
Total 1 6 6 8
Thus 84% of the students have passed the exam.

The laboratory exercise results:
G
D 2
Dr 1
FK 1
IM 17
M 1
Total 23
92% of the students have passed the laboratory exercise.

Teaching

We have tried to combine theoretical lectures with immediately following hands-on exercises on the just covered material. To encourage work on these exercises the students were given bonus points on the exam if they were demonstrated before a given deadline.

Examination

The examination was both an ungraded laboratory exercise and a written exam.

The laboratory exercise was performed in groups of two students and then presented both as a written report and a demonstration of the developed visualisation (of airflow in two different rooms).

The written exam was based on a few questions on terminology and theory and the rest on practical application of visualisation methods on various example data.

Course evaluation

No course evaluation questionnaire was distributed, so the following points are based on informal discussions with students and my own perceptions.

The most important change in this year's course was that the VTK programming library was dropped and instead all exercises done in OpenDX. This definitely simplified matters for the students, in fact to the point that they probably spent too little time on the final laboratory exercises, in most cases choosing the automatic visualisation created by OpenDX, rather than thinking through what they should visualise. The exercise will have to be modified in order to make it more clear what level of effort is needed.

In the light of the requirements to save time and money, some things could be changed:
The oral presentation of the exercise could be skipped. However, in this case it would be preferrable to let the students do the exercise singly, rather than in pairs, but this on the other hand will require more workstations.

While OpenDX was easy to learn and use, it still required some amount of hands-on demonstration, so it would be difficult to have unsupervised lab sessions. It would perhaps save some time if the exercises were not counted towards exam bonus and therefore did not have to be presented. Hopefully the students would do them anyway.

The time spent on the course did not seem to be excessive and most of the exercises could be done in the time allotted, a bottleneck perhaps being a lack of lab assistants.

Dropping VTK also meant dropping the previous course book. The OpenDX book was intended as course literature instead, but the student apparently felt it to be too expensive so only one person actually bought it. A number of articles were also supplied as course literature, but while these hopefully gave insight into specific visualisation methods, there is still a lack of a book giving a good presentation of the process of visualisation. The hunt for a good book goes on.

The theory material on some of the lectures could probably be expanded somewhat, so there's opportunity for improvement for next year.

In summary, I have managed to implement most of the plans from last year, except finding a good course book.