Login  Register

Re: [fiji-devel] Re: Google Summer of Code

Posted by dscho on Mar 29, 2009; 9:48am
URL: http://imagej.273.s1.nabble.com/Google-Summer-of-Code-tp3693141p3693151.html

Hi Gabriel,

On Sun, 29 Mar 2009, Gabriel Landini wrote:

> On Sunday 29 March 2009, Johannes Schindelin wrote in the IJ list:
> > Oh, and if you have an idea for a project that is so obviously Fiji
> > that we cannot possibly reject it, but which is not listed on the Fiji
> > Wiki: the ideas we listed are only suggestions, food for thought.  Go
> > wild.
>
> There are a number of mini-projects/ideas in the ImageJ wiki site (the
> Wishlikst http://imagejdocu.tudor.lu/doku.php?id=wishlist:start) which I
> have been wondering whether they would be suitable.

Hmm. If I can choose between students that implement ImageJ plugins, and
students that work on the core infrastructure of Fiji, I will pick the
latter.

Of course, it would be nice to have all these plugins implemented, but
when do you get the chance to work on a bigger project for 3 months
straight?

> I am adding some more here:
>
> 1) Alpha_shapes / concave_hull (could be part of item 7 below)
>
> 2) Bilateral Filtering

Is in Fiji already.  That was a really small project.

> 3) Colorizing algorithms (i.e greyscale to colour,there are several
>    methods)
>
> 4) Retinex algorithm
>
> 5) Interactively adjustable intensity/LUT curves (not just linear
>    controls, but splines/etc based)
>
> 6) Support for storing ROIs in TIFF tag fields, so ROIs can be saved
>    with the images).
>
> 7) Graph Theory constructs (in addition to the Delaunay/Voronoi)
>
> 8) Expand the ImageFlow module base (not sure if the source is available
>    yet).

The source is not available yet, but I talked to Daniel about it already.  
He will be fine to have the source code in Fiji, once he released the
source to the public.

> 9) Object (rather than image) selector/sorter (something like Kai's
>    pixolu.de If you haven't tried it, you should) to create "objects
>    like these"  filters/classifiers to be extracted from images and sets
>    of images.
>
> 10) Interface between R and ImageJ/Fiji (Dimiter has done something like
>     this, but it would be nice to have a set of implemented procedures
>     so IJ/Fiji can run statistical procedures directly from Results
>     tables, etc)

R is platform specific.  As we support 7 platforms already (and somebody
is working on an 8th), that would add a tremendous maintenance burden.

I'd rather have an existing Open Source statistical package written in
Java integrated.

> 11) A set of more powerful painting brushes and image editing tools.
>
> 12) Cross platform webcam support (to produce cheap
>     videomicroscope/telescope units)

Again, this is very platform specific, what with JMF not living up to its
promises.

>
> 13) Statistical region merging segmentation:  
>     http://www.univ-ag.fr/~rnock/Articles/Drafts/tpami04-nn.pdf

I am in the middle of writing that already (would be finished if I had not
gotten side-tracked yesterday).

> 14) Fractal interpolation resizing (from the GIMP GSoC entry)
>
> 15) A unique/common segmentation interface (I have collected near 15 new
>     histogram segmentation methods that would be better put under a
>     single interface together with others already available).
>
> 16) A javascript Recorder (similar to the Macro Recorder but producing
>     javascript instead).

Why limit it to Javascript?

> 17) Virtual microscope-like image viewer.
>
> 18) HSB/Lab painting modes
>
> Is that wild enough? :-) I am sure others will have further ideas.

Heh.

Most of these ideas are too small to make even half a 3-month project,
would you not agree?

And I see only 10) 12) and 16) being Fiji-related enough for a Fiji
project, no?

Having said that, I find all of your ideas interesting!

Ciao,
Dscho