MATLAB Central Newsreader - impoint slow with on large figure/image

impoint slow with on large figure/image

Abe Lau — Thu, 23 Oct 2008 00:50:04 -0400

Dear all,
I am trying to use the function "impoint" to create user draggable points where some calculation is based upon on an GUI. However, I found that with a large image, the selection of point is not very responsive, with the first left click end up dragging the point instead of setting it initially.

e.g.

test = zeros(3000,3000);
figure; imshow(test, []);
pt = impoint;

Is it an expected behavior, or is there any workaround for it?
Thanks,
Abe

Re: impoint slow with on large figure/image

Patrick R — Wed, 10 Jun 2009 10:04:01 -0400

I experience the same weird behavior with impoint and tried to fix it now for days. Unfortunately my matlab knowledge is not advanced enough to mess around in the default functions of the image processing toolbox, but I really go crazy with this problem.

in fact, I just need to let the user define one point in an image to start a semi-automatic-segmentation from.
the function getpts() would also work well enough for me, except for the fact that you can't restrict it to one single point but letting the user choose as many as he likes to.

has no one a fix to that problem? i work with matlab r2009a.
thanks in advance,
patrick

"Abe Lau" <abe@nonexistancegmail.com> wrote in message <gdohns$foe$1@fred.mathworks.com>...
> Dear all,
> I am trying to use the function "impoint" to create user draggable points where some calculation is based upon on an GUI. However, I found that with a large image, the selection of point is not very responsive, with the first left click end up dragging the point instead of setting it initially.
>
> e.g.
>
> test = zeros(3000,3000);
> figure; imshow(test, []);
> pt = impoint;
>
> Is it an expected behavior, or is there any workaround for it?
> Thanks,
> Abe

Re: impoint slow with on large figure/image

Steven Lord — Wed, 10 Jun 2009 14:12:21 -0400

"Patrick R" <p.reichmann@inode.at> wrote in message
news:h0o0eh$r6c$1@fred.mathworks.com...
>I experience the same weird behavior with impoint and tried to fix it now
>for days. Unfortunately my matlab knowledge is not advanced enough to mess
>around in the default functions of the image processing toolbox, but I
>really go crazy with this problem.
>
> in fact, I just need to let the user define one point in an image to start
> a semi-automatic-segmentation from.
> the function getpts() would also work well enough for me, except for the
> fact that you can't restrict it to one single point but letting the user
> choose as many as he likes to.
>
> has no one a fix to that problem? i work with matlab r2009a.

For your application, you may be able to use GINPUT instead of IMPOINT.

--
Steve Lord
slord@mathworks.com

Re: impoint slow with on large figure/image

Patrick R — Wed, 10 Jun 2009 14:31:03 -0400

"Steven Lord" <slord@mathworks.com> wrote in message <h0oev9$s6m$1@fred.mathworks.com>...
>
> "Patrick R" <p.reichmann@inode.at> wrote in message
> news:h0o0eh$r6c$1@fred.mathworks.com...
> >I experience the same weird behavior with impoint and tried to fix it now
> >for days. Unfortunately my matlab knowledge is not advanced enough to mess
> >around in the default functions of the image processing toolbox, but I
> >really go crazy with this problem.
> >
> > in fact, I just need to let the user define one point in an image to start
> > a semi-automatic-segmentation from.
> > the function getpts() would also work well enough for me, except for the
> > fact that you can't restrict it to one single point but letting the user
> > choose as many as he likes to.
> >
> > has no one a fix to that problem? i work with matlab r2009a.
>
> For your application, you may be able to use GINPUT instead of IMPOINT.
>
> --
> Steve Lord
> slord@mathworks.com
>

thanks alot Steve, I will try that right away.

Re: impoint slow with on large figure/image

Patrick R — Thu, 11 Jun 2009 10:08:01 -0400

"Patrick R" <p.reichmann@inode.at> wrote in message <h0og37$fj5$1@fred.mathworks.com>...
> "Steven Lord" <slord@mathworks.com> wrote in message <h0oev9$s6m$1@fred.mathworks.com>...
> >
> > "Patrick R" <p.reichmann@inode.at> wrote in message
> > news:h0o0eh$r6c$1@fred.mathworks.com...
> > >I experience the same weird behavior with impoint and tried to fix it now
> > >for days. Unfortunately my matlab knowledge is not advanced enough to mess
> > >around in the default functions of the image processing toolbox, but I
> > >really go crazy with this problem.
> > >
> > > in fact, I just need to let the user define one point in an image to start
> > > a semi-automatic-segmentation from.
> > > the function getpts() would also work well enough for me, except for the
> > > fact that you can't restrict it to one single point but letting the user
> > > choose as many as he likes to.
> > >
> > > has no one a fix to that problem? i work with matlab r2009a.
> >
> > For your application, you may be able to use GINPUT instead of IMPOINT.
> >
> > --
> > Steve Lord
> > slord@mathworks.com
> >
>
>
> thanks alot Steve, I will try that right away.

GINPUT works perfectly fine, thank you very much Steve, that one really helped me a lot.
cheers, Patrick

Re: impoint slow with on large figure/image

Alex Taylor — Thu, 11 Jun 2009 20:27:02 -0400

"Abe Lau" <abe@nonexistancegmail.com> wrote in message <gdohns$foe$1@fred.mathworks.com>...
> Dear all,
> I am trying to use the function "impoint" to create user draggable points where some calculation is based upon on an GUI. However, I found that with a large image, the selection of point is not very responsive, with the first left click end up dragging the point instead of setting it initially.
>
> e.g.
>
> test = zeros(3000,3000);
> figure; imshow(test, []);
> pt = impoint;
>
> Is it an expected behavior, or is there any workaround for it?
> Thanks,
> Abe

Abe,

Thank you for reporting this issue. This is a bug that we are addressing in a future release.

Thanks,

Alex.

Re: impoint slow with on large figure/image

Abe Lau — Thu, 11 Jun 2009 23:34:02 -0400

"Alex Taylor" <alex.taylorREMOVE.THIS@mathworks.com> wrote in message <h0rpam$es5$1@fred.mathworks.com>...
> "Abe Lau" <abe@nonexistancegmail.com> wrote in message <gdohns$foe$1@fred.mathworks.com>...
> > Dear all,
> > I am trying to use the function "impoint" to create user draggable points where some calculation is based upon on an GUI. However, I found that with a large image, the selection of point is not very responsive, with the first left click end up dragging the point instead of setting it initially.
> >
> > e.g.
> >
> > test = zeros(3000,3000);
> > figure; imshow(test, []);
> > pt = impoint;
> >
> > Is it an expected behavior, or is there any workaround for it?
> > Thanks,
> > Abe
>
> Abe,
>
> Thank you for reporting this issue. This is a bug that we are addressing in a future release.
>
> Thanks,
>
> Alex.

No worries. really glad to hear an update on this.
Just a side question.... is there any kind of bug database we may possibly make reference to/search ... rather than waiting for developers to fish for possible problem? It's great to know that someone pick it up, but it doesn't look very efficient.

Abe

Re: impoint slow with on large figure/image

Steve Eddins — Fri, 12 Jun 2009 11:15:05 -0400

Abe Lau wrote:
>
> [snip]
>
> No worries. really glad to hear an update on this.
> Just a side question.... is there any kind of bug database we may possibly make reference to/search ... rather than waiting for developers to fish for possible problem? It's great to know that someone pick it up, but it doesn't look very efficient.
>
> Abe

The most efficient way to view and report bugs is via this page:

http://www.mathworks.com/support/

---
Steve Eddins
http://blogs.mathworks.com/steve/

Re: impoint slow with on large figure/image

Abe Lau — Fri, 12 Jun 2009 11:45:02 -0400

Steve Eddins <Steve.Eddins@mathworks.com> wrote in message <h0tdbp$4v2$1@fred.mathworks.com>...
> Abe Lau wrote:
> >
> > [snip]
> >
> > No worries. really glad to hear an update on this.
> > Just a side question.... is there any kind of bug database we may possibly make reference to/search ... rather than waiting for developers to fish for possible problem? It's great to know that someone pick it up, but it doesn't look very efficient.
> >
> > Abe
>
> The most efficient way to view and report bugs is via this page:
>
> http://www.mathworks.com/support/
>
> ---
> Steve Eddins
> http://blogs.mathworks.com/steve/

Thanks Steve. I really completely missed that.... I remember looking for bug report database few years ago but never find it haha.