DSP Robotics Support

I'm thinking to start something like this project:
http://www.bion.si/DKF_E.htm
http://www.bion.si/DVB/DVB_E.htm
Have some PDFs with additional math explanation for that, so upload them later.
Do you think Flowstone would be good for it?

Very interesting, tester.
It is always good to find an example of real scientific exploration so accessible to the "amateur" - the tools can be cheap enough so that one does not need to seek approval from the established formal bodies with their dogmatic ideas of what is good to study.

From a technical sense...
Yes, I think even with the standard FS tools this could be done - it would not be very fast, I think, but in this case the accuracy of unmasking the small fluctuations from the camera shot noise is the primary requirement.

The R,G,B colour information for a pixel can be found using the 'Get Pixel' primitive for any loaded image, and the "Set Pixel" could be used to construct the output bitmap - thus making a crude "array reader" and "array writer" for the pixel colour information which can be scanned using integer loops.
Here's a quick example that creates a btmap which shows the colour difference between two source bitmaps...

For other processes, only the "PROCESS" module would require changing - though it would become quite a complex module when it becomes necessary to extract the values of surrounding pixels - e.g. for creating blur or edge detection.
I tried this with some larger bitmaps - approx.2000*3000px - and it seems to work fine, though as predicted, it is rather slow (about 1 minute per image on my little laptop).

(PS - don't worry about the error boxes at startup; the stock bitmap display doesn't like empty bitmaps - the error will go away when you load something.)

This one could be ruby based I guess. I'm not mixing ruby into my audio stuff in order to avoid unpredictable consequences on the finish line, but this one perhaps could be done with the help of ruby.

Here are promised additional documents.

I think I again contact that fellow who makes this research, and maybe he is willing to give more specs on his software/procedures. As far I remember, his app wasn't finished nor stable (I have some unstable exe here from that project), but it wasn't FS and it required more programming skills. Made in FS - could be easier and better, and more accessible to audience. Would be a shame if this project went forgotten due to lack of money or interest in developing it further.

tester wrote:This one could be ruby based I guess.

Possibly - tricky thing is that in Ruby there's no access to the pixels inside the bitmap at the moment - all you can do with bitmaps is draw them to the screen.
But, for the array manipulation, Ruby would be well suited - and has better mathematical precision too.
I've had some success loading the raw binary data from .wav files in Ruby - so, for uncompressed bitmaps at least, it may be possible to do the same thing. If I get a moment, I'll see if I can find some file spec's and look into that.

Had a quick look at the documents - much to digest, but the kind of processing being done to the images seems fairly straightforward. I've been keen on digital photography for a long time, so I already have some familiarity with the different noise sources - including using averaging and dark frame subtraction in the context of noise reduction.

I currently use a piece of software called Combine ZP for these tasks - designed for the processing and analysis of microscope images.
It is not the the easiest thing to use, but it allows batch processing of any number of files for averaging, filtering, automatic alignment etc. - maybe it will be useful to you, or inspire some ideas for how your FS project should function.

The piece of software I have from the researcher - does the video processing (but is painfully unstable). Hopefully FS will offer more tools for video stream (and thus - graphics) manipulations, some modules are built in I saw. It's still DSP, just in different area.