student research project and krita

[Boudewijn Rempt]

On Thursday 23 February 2006 16:04, Florian Merz wrote:
> Hi everybody,
>
> I have to do some kind of research project/seminal paper for university
> (Studienarbeit in german) and I thought maybe I can use this to get into
> krita development if I can find a suitable subject.

Good! There should be plenty of stuff to do. 

There's the natural media stuff -- inventing good paint, brush and rendering 
models. Bill Baxter's work is very important here, but it's a field with 
quite a bit of history, from Hairy Brushes to Wet & Sticky.

I've been thinking that it would be nice to have RAW colorspace -- that means 
you don't convert the RAW images from your camera on loading, but just load 
the raw and edit it. Would be a bit slow, but very innovative. Real time 
rendering of color-corrected raw, that would be very exciting. 

Color mixing according to the laws of nature. Creating pleasing palettes from 
existing images.

A way to split an image into its component objects: take a picture of three 
people, run the splitter and end up with four layers: one background and one 
layer per person.

Porting lcms to openGL shaders: that would speed up color correcting, would be 
very innovative, widely useful and you might get money from NVidia or ATI to 
do that. (Although nvidia probably wants you to use Cg, which is a bummer).

Using opengl brushes -- but that would tie in with the first idea, and Art 
Rage already does that. Maybe not innovative enough, but exciting anyway.


> The project should be worth about three months of work but I guess about
> half of the time is needed for preparations, writing the paper and creating
> a presentation so it would probably sum up to about one and a half month of
> coding work. As far as I know most students need more time than those three
> months (some even up to a year) so this is not exactly a hard limit.
>
> I'm writing to this list because I wanted to ask if anyone of you knows
> about a suitable subject to write about which might result in an
> interesting feature for krita. The subject should have to do with image
> processing and be of some theoretical/research interest otherwise it
> probably won't get accepted. This is not a thesis so it doesn't have to be
> something completely new but some new aspects (to some other research
> project/paper) would be nice to have.

Doing a _good_ chalk simulation. Dust, height fields, stumping.

> I know that the krita team is especially interested in the simulation of
> real painting but I don't know if anything in this direction would be of
> interest for my prof so I would be glad for some other suggestions, too.
>
> Regards,
>  Florian
>
> p.s: I'm not afraid of math and some impressive formulas (for the
> presentation) are always welcome ;)

Have you read Bill Baxter's thesis? It's full of formulas, and I don't have 
any math skills!

-- 
Boudewijn Rempt 
http://www.valdyas.org/fading/index.cgi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://mail.kde.org/pipermail/kimageshop/attachments/20060223/af0b59e0/attachment.pgp