[Scilab-users] ?==?utf-8?q? ?==?utf-8?q? ?= creating ROI from mouse click

[Antoine Monmayrant]

Le Vendredi 11 Mars 2016 10:11 CET, Philipp Mühlmann <p.muehlmann at gmail.com> a écrit: 
 
> Thanks, but wouldn't this limit the ROI shape to be a rectangle?

Yes, sorry, I did not understand your requirements.

> 
> My purpose would be to have a ROI of any possible shape one can get with
> mouse clicks, having the mouse positions "connected" via straight lines.
> 
> I'd like to have a look inside of "rubberbox.sci" or even "xfpoly.sci" to
> maybe adjust it to my needs...but I can't find it.

You can't. It's a hard-coded function.
What you can find in the source code is a java function that does the rubberbox, similar to the one that controls the zoom.

Concerning xclick, I can't figure out the meaning and use of the "flag" input parameter...

Antoine

> 
> There is also an imroi-function coming with SIP-Toolbox.
> But this also deals only with rectangles and ellipses
> 
> BR
> Philipp
> 
> 
> 2016-03-11 8:59 GMT+01:00 Antoine Monmayrant <amonmayr at laas.fr>:
> 
> >
> > Le Vendredi 11 Mars 2016 06:39 CET, Philipp Mühlmann <
> > p.muehlmann at gmail.com> a écrit:
> >
> > > Dear all,
> > >
> > > how can I use mouse clicks to create a region of interest (ROI) for image
> > > processing without using IPD toolbox?
> > >
> > >
> > > My current aproach is something like this:
> > >
> > >
> > >
> > > 1.) display an 2-D array, e.g.: a grayscale image
> > >
> > > 2.) use xclick to get mouse coordinates
> > >
> > > 3.) use xfpoly to draw a mask image based on the points clicked with the
> > > mouse
> > >      --> mask image has same size as 2D-array in 1.)
> > >      --> mask image starts as zeros only
> > >      --> xfpoly will fill the polygons white =ROI
> > >      --> result is a 2D array containing only "0" and "1" --> can be
> > saved
> > > as b/w image
> > >
> > > 4.) use xs2bmp to save the mask image temporarily
> > >
> > > 5.) use "ReadImage" from IPD-toolbox to load the temporary image and
> > delete
> > > this temporary image afterwards
> > >
> > > 6.) now the mask can be applied onto the original image such as:
> > > result_image = input_image .* mask_image;
> > >
> > >
> > >
> > > I'd prefer to get already the position of the area filled by xfpoly in
> > step
> > > 3.) as a variable.
> > > This would prevent the need of the temporary image and the need of using
> > > IPD.
> >
> > Hi,
> >
> > I am not sure this corresponds to what you are looking for, but an
> > alternative to xclick could be rubberbox:
> >
> > [final_rect, btn] = rubberbox()
> >
> >
> > Antoine
> >
> >
> > >
> > > Any idea?
> > >
> > > BR
> > > Philipp
> > >
> > > --
> > > There we have the salad.
> >
> >
> >
> >
> >
> >
> > _______________________________________________
> > users mailing list
> > users at lists.scilab.org
> > http://lists.scilab.org/mailman/listinfo/users
> >
> 
> 
> 
> -- 
> There we have the salad.