More information about the individual components will be detailed later. See the right hand menu.
Due to many requests, and because it will not delay the marking, the Portfolio deadline has been postponed until Monday 11 May 12noon. I know that this is an hour after your exam. If this is a problem, please choose your own deadline at some point on Friday 8th.
I have simplified the exercises for Week 7, so that you should be able to do the exercises which are now presented. Please have a go, and let me know if it is still unreasonable difficult.
An example exam question has been publish. See exam in the right hand menu. You can also find the raw images used in that question. You may want to use them for comparison when you debug your own chi2 test.
Information about the poster project has now been posted. Note in particular that I expect you to form teams of 3-5 by 17 February.
I want to give you as much time as possible to discuss the topic and contents of the poster. Therefore, students who have not teamed up by 17/2 will be assigned randomly to groups.
Given that we have now lost one lecture due to the nations inexperience with snow, my plan is to delay subsequent lectures by a week. I am trying at the moment to get a lab for Week 7, to avoid losing lab time.
The ninth and last lecture would then be given in Week 10. There will be no discussion/peer assessment of Exercise 9, as I still want to give you the portfolio assignment before Easter.
This plan is assuming that I get a lab for Week 7.
I got a question about the extract function. The following text has also been added to the new version of the lab sheet.
If you do the extract function in the most generic way,
it does not know how long the message is. The ideal implementation
would take all pixels as message (discarding odd pixels at the end
which don't make up a full character), and convert it to a string.
Done correctly, this should give you the message first and then
rubbish at the end (from random-ish LSB-s from unused pixels).
There are possible variations.
You may give extract an extra parameter containing the
message length. This is ok, but not as good because Bob will
not always know the length.
It is also possible to store the message length as part of the
message (using, say, the first 16 bits embedded to store an
integer giving the number of characters embedded).
Please note that the lab sessions will be in 34(a)BB04. It seems that we only had the small lab on BB03 booked (18 seats). In BB04 we have 32 seats which is likely to suffice.
I got some questions about running Matlab on Windows/at home/elsewhere, and I did not have complete answers at the time. It seems that this is harder than I thought it was, but there are some options.
Matlab is available on Windows, Mac, Solaris, Linux, et c., but it is rather expensive software; the engin may be cheap, but the toolboxes are expensive. You need the image processing toolbox. There is a student version which is cheaper than the full version, but this does not necessarily mean cheap. I don't know of anyone who uses it.
You can run Matlab remotely off the faculty unix system; whether from home or from Austin Pierce. The GUI windows are only practical when you have a fast network. If you want to do this from Windows, you need an ssh client and an Xserver; see the software page (right hand menu) for details.
On most unices (incl. linux) you already have an Xserver,
and ssh should be installed; run
ssh user@student.ee.surrey.ac.uk, and you will
get a new prompt from the faculty box.
Everything should work as if it were local, except that
GUI windows are a pain over a slow network...
$Id: index.php 1499 2009-05-05 10:36:36Z css1hs $