The final application could read in an 8-bit greyscale image (as iris recognition devices usually take infra-red greyscale pictures), make an attempt at auto-detecting the pupil, iris and eyelids. Then it will encode the iris into a database, and allow you to compare other irides to it in the future.

I was a little weary at first about doing a project on image processing, but it’s been very fun and interesting to see how different algorithms work for i.e. edge-detecting (sobel filter), blurring (median filter), etc.

Early on we had the choice of using Java or C++ for the application. We decided to go with C++ and Qt and it really paid off. Our entire Qt experience was painless and filled with pleasant surprises along the way of all the Qt classes available to us. QPixmap, QImage and even QtImageFilters are actually extremely powerful and get the job done. We were constantly impressed by the clean solutions that Qt had for everything.

We also tried it briefly on Windows and Mac along the way so the application works well for those OSs.

Found along the way

Some interesting facts I found out along the way:

• The pupil and the iris are not concentric (vary slightly)
• Iris patterns remain remarkably stable over your lifetime (and do not depend on your eye colour)
• This depends on your parameter, but the chances of a false match can be as good as 1 in 1.5 billion
• We implemented a median filter that has order O(r)
• Finally, an interesting story

The software is available under the GPL licence, though it will probably only be useful for anyone who has a small software project similar to this one to do

Source code and report

Our paper for the project and the source code is available at:
projectiris.co.uk

They include: a commenting plugin, a buffer explorer plugin, a yankring plugin, better prolog/Qt syntax highlighting, easy .h -> .cpp switch plugin, some others, and a lot of shortcuts (like for pasting from other applications). Also the very essential tetris plugin.

Vim Awesomeness

Throughout <C-X> means Ctrl-X

First get the config — just execute this command and it will do it:

cd && mv .vimrc{,_old} && mv .vim{,_old} && wget http://www.doc.ic.ac.uk/~kg109/configs/vim{.tar,rc} && tar xf vim.tar && mv {,.}vimrc

Read on for the cheat sheet…

Navigation:

Go back to normal mode — <Esc> or easier: <C-c> or <C-[>
Move Around — Down (j), Up (k), left (h), right (l)
Move word — w (word forward), b (word backward) ; W (word forward — skip punctuation), B (word backward — skip punctuation)

Page Jump — Page Down (<C-f> — forward), Page Up (<C-b> — backward)
Page half jump — Down (<C-d> — down), Up (<C-u>)

End of Line — $
Start of Line — 0

Start of document — gg
End of Document — G

Jump to last place that you made an edit — g; (g semi-colon — you can also keep pressing it to keep going back)
Jump forward/back paragraph (next line with space) — { or }

Visual Mode:

Enter visual mode — v
Visual line mode — V
Visual block mode — <C-v>

Find-and-replace over some selection — select it in visual mode, then type: :s/oldword/newword/g after you type the first : it will expand to :’<,’> which is fine

Insert Mode:

Insert mode from before the cursor — i
Insert mode from after the cursor — a

Jump to the end of the line and go into insert mode — I (capital i)
Jump to the beginning and go into insert mode — A

Make a new line below and go into insert — o (line above is O)
Auto-complete variables/functions etc — start typing it and then <C-n> or <C-p> next and previous

Deleting:

delete character — x
delete line – dd
delete from cursor to end of line — D
delete several lines — select them in visual mode, and then d or x

delete whole document — ggdG — which does gg (start of document), d (delete) ..G (end of document)
Delete word – dw or dW (word with punctuation)

Delete everything inside {} braces — diB — delete inner brace
Delete everything inside the {} and the {}s — daB — delete a brace

Delete everything inside () brackets — dib — delete inner brackets
– dab — delete a bracket and its contents

Jump to equivalent bracket / brace — % — so if you’re at a { type % and it goes to the relevant }

Clearing

Clearing is just like deleting, but it puts you in insert mode after the deletion. You can also always use it in any case where you use d, so for example:

cw — clear word
cc — clear line
C — clear until the end of the line
ciB — clear inner brace

……… and all the other ones from the Deleting section

Auto-indent

Auto-indent line — ==
Auto-indent whole document — gg=G which goes to start of doc, and indent until the end

Editing files:
Open and edit a new file: from normal mode, type: :e filename.cpp

Find / Replacing

From normal mode:

/something — searches for something. Then ‘n’ searches for the next term forward, and ‘N’ backward
find and replace — :%s/oldword/newword/g
find and replace in all open files —        :bufdo %s/old-word /new-word /ge | update

You don’t have to use the / character by the way — you can use any other character like , or ~ So for example to find-and-replace you could:

:%s~old-word ~new-word~g

Buffers / Windows

Next buffer — <C-j>
Previous Buffer — <C-k>
Close a buffer: :bd — buffer delete

Split windows: :split
Vertical split: :vs — :q to close any of those (won’t close the buffer — :bd to do that)

Cycle through windows: <C-ww>
Switch between .h -> .cpp — <C-a>

Saving / Closing:

Save file: :w
Quit: :q or quit-without-saving —   :q!
Save and quit —   :wq or easier:    

Save all open buffers: :wa

Copying / Pasting:

A delete/clear in vim is a ‘Cut’ — so you can always paste it after. Do just ‘copy’ or ‘yank ‘:

yank line – yy
yank word –yw
yank to end of line — Y

…. you get the idea

Cycle through paste history — paste something (p) and then hit <C-p> to go back in history, or <C-n> to go forward. Alternatively, just hit <F10>

Commenting:

Comment out a line — – (literally the dash character)

Comment out “sexy style” : ,cs (comma c s) i.e. it will do a comment like:

/*
* Line commented out
*/

Switch comment style — ,ca — i.e. switch between using // or /* in C++

Repeating Actions:

Vim is uber consistent. So for most of the above, you can do them X amount of times by proceeding the command with X. So for example:

Delete the next five words: 5dw
Yank the next five lines: 5yy

….etc

Repeat last action: . (full stop)

Macros

Think of them as VBA macros — vim will record the key-presses you make, and when you call the macro it will execute them again:

You can record a macro and place it under any character name. For exactly, to place it under d, just type:

qd <perform all mouse movements > then hit q
Now you can call that macro again with: @d
Execute it four times for example with: 4@d

Folding

For large files, you might want to fold/hide away certain functions temporarily.
Fold selection — select it in visual, then zf
To open a fold: zo
Not a fold, but will wrap the line to a certain length — gq

Extra other configs:

Toggle line numbers off / on: <F12>
Look at paste history and paste something from it: <F10> then navigate through and <Enter>

Tetris:

:Tetris

KDE 4 Reloaded Live CDs / USBs

Firstly, the Build Service recently acquired a functionality to generate your own system images from the build service, and the KDE guys quickly generated openSUSE 11.1 KDE4 “Reloaded” CDs and USB images containing KDE 4.3.1. That means they include all security etc. updates (from the openSUSE update repository) as well as the latest KDE 4 packages, and Firefox 3.5.

SUSE Studio

SUSE Studio was fully launched, with quite literally the most usable interface you can imagine for generating any kind of appliance that you want. Great to play around with, but also very useful to have your fully customised USB image, filled with built-in pictures of yourself to flatter your ego. One of the things that impressed me the most was the test drive service, which lets you try out the appliance in a web browser before downloading it.

• Logging the meeting and automatically converting it to a more readable HTML file

• Automatically generating minutes by listening to specially selected commands (click the link to see an example)

Since writing minutes after a meeting can sometimes be an onerous task, this should really help in that department.

If you like, compare the commands in the log (all beginning with # — i.e. #action something) to what is output in the minutes to get an idea of what things can look like.

It was created some time ago and has been used in-and-around Debian, Ubuntu and Fedora project channels, but I don’t think it’s been public knowledge everywhere just yet. Anyone can use it straight away from any channels that bugbot is in; any problems, please give me a shout!

Thanks to Richard Darst and all the others involved in the creation and development of the supybot plugin.

“President Obama has said he is prepared to go through the budget “line by line” to cut wasteful spending…”

McNealey, asked to write a report on the advantages, comments:

“It’s intuitively obvious open source is more cost effective and productive than proprietary software,” …”Open source does not require you to pay a penny to Microsoft or IBM or Oracle or any proprietary vendor any money.”

The article also contains quotes from OSI President Michael Tiemann:

“It’s an accident of history that proprietary standards became so entrenched so early and it’s been a colossal expense for government.”Mr Tiemann said while some departments already use open source technologies, overall it has been estimated that the global loss due to proprietary software is “in excess of $1 trillion a year.”

Here’s to hoping we see some action followed by McNealy’s report.

For a preview of how this happens, join #kde-bugs-dist. Basically, when there’s a new bug, bugbot will drop in a note with:

Of course, bugbot will still snarf out bugs, i.e.:

So if you have any KDE project IRC channel, and you want bugbot to snarf bugs and/or report new ones in your channel, just drop me a note (apokryphos in #kde-devel, #kde-bugs) and I’ll have him added.

Continuous thanks goes to Mez for his gracious hosting.

So far I’ve been really impressed with the latest release. Not only have the reviews been pretty much consistently positive because of some of the shiny new features (like installer and fast package management), but also in general it’s just running very well. I installed it on some 10 different desktop and laptop machines over the past few weeks, and every single one has gone very smoothly.

KDE 4.1

There are a lot of nice things in the upcoming 4.1, so I’ll just briefly mention a few of those visual ones which particularly interest me:

There have been several additions to the KWin composite effects including wobbly windows:

And some new window switching effects:

 

A desktop cube plugin is also in the works.

The Oxygen icon theme is also a real pleasure to use (coming in different colours):

I’ve been following the development version (with of course KDE 4.0) for some time now and I’m really pleased about the way it has turned out! Distrowatch and others have been giving some glowing reviews, and the general feel from testers of the release appears to be incredibly positive. The new installer, sexy artwork, super-fast package management, and new official KDE 4 live CD is really making a big change. So congratulations to everyone who made it happen!

Packaging With the openSUSE Build Service II

When: Friday, April 4th, 2008 through Saturday, April 5th, 2008 in every timezone.
Where: #opensuse-buildservice

Another good opportunity to get to grips with the openSUSE Build Service, the easiest way to package your applications for many distributions, in one place, and — thanks to AMD — on a build host of over 100 CPUs. Currently supported distributions are:

• CentOS 5
• Debian Etch
• Fedora 7
• Fedora 8
• Fedora Core 6
• Mandriva 2006
• Mandriva 2007
• Mandriva 2008
• RHEL 4
• RHEL 5
• SLES 9
• SLE 10
• SUSE Linux 10.1
• openSUSE 10.2
• openSUSE 10.3
• openSUSE Factory
• Ubuntu 6.06
• Ubuntu 7.04
• Ubuntu 7.10

With the build service you can also generate YMP files used for 1-Click-Install, which has been supported for some time on openSUSE, although a client has already also been created for Ubuntu and Debian.

There will be many KDE, GNOME and build service celebrities hackers around to help with any questions you might have. Getting a build service account is instant, and the software itself is of course also free software licensed under the GPL. There is a pretty web interface:

…and a command-line client, so do drop by on the day if you need any help while playing around with either of these.

KDE Bug Day: Konqueror Bugs

When: 0:00 UTC – 23:59 UTC on April 6th 2008
Where: #kde-bugs

Another bug squashing day taking place in #kde-bugs this Sunday; the first one in a long time. The product concentration will be Konqueror bugs, and bugbot will of course be around all day to help. The TechBase page lists all the information. Make sure you’re there!

Very glad to see successful KDE (with Amarok) and openSUSE booths about. We gave out over a whoopin’ 1000 openSUSE Promo DVDs (Live+Install, with KDE/GNOME/Xfce).

Our KDE 4 desktop received a lot of attention, and unfortunately quite a few crashes at some points as we were running a very recent trunk snapshot. Still, many users very impressed with Dolphin, the new Kate, Gwenview, KWin Composite effects, the Kickoff menu and our super-sized Plasma clock. To put your mind at ease, I can promise you that we have no direct affiliation with Flavor Flave.

Some things in particular that I took from the event:

Linkat: a distribution for the education department in Catalunya

The Jordis presented the very professional Linkat, a SUSE-based Linux distribution now endorsed and promoted by the Catalynian government for schools and communities there, with all kinds of FOSS educational software.

Interesting to hear about the many challenges faced: for example, they were asked to include the latest packages but to also provide long term support. So what did they do? They use openSUSE packages when they need to have the latest-and-greatest (which comes with 2 years of security and update support), and SLED for the base packages and everything else (7 years of support). Watch this space.

The new Sat Solver

Probably the talk that personally interested me the most. Michael gave a quick overview of the workings into making a package management solver. The new one is completely based on standard SAT algorithms (boolean satisfiability problem), which has nice advantages like the simple fact that there are insanely fast algorithms for computing such problems. Such small problems (in comparison) like our “very complex” package management decisions take milliseconds under this model.

This, combined with the new solv files (for repo metadata), makes for extremely fast package management.

Other highlights:

• Timo’s very well-attended Kernel, udev, D-Bus, HAL, NetworkManager and Friends gave a really nice and quick insight into how you can debug any problems with this stack using his tidy little event notification application. The example provided was unsupported multimedia keys (surprisingly easy to fix). Will find the web link for it soon.
• Nice talk on KIWI which I’ve written about before. The more time that passes, the more I’m convinced that applications designed as distribution-independent tools (like the OBS, PackageKit, Klik, Smolt) like KIWI are completely the way forward.
• Great to finally properly meet so many more developers (too many to list).

Huge thanks to Pascal Bleser and the rest of the FOSDEM team working tirelessly to ensure that FOSDEM ran so smoothly. Nice events like that make me very happy; looking forward to FOSDEM 2009!