Well, I used pix_blob, or blob tracker, do not remember exactly, but I always gave erratic readings or based only on two axes, I like working with coordinate and mgrid are well suited to that need. Mathieu Gridflow ever asked me, I have not tried what Derek and Mathieu proposes no doubt I will today, but obviously I ask the question because I do not like to rely on a single object or platform.
Best regards friends
José
lets make a replacement patch. where shall we begin?
pp
________________________________________Sent: Tuesday, December 07, 2010 4:32 PM
From: pd-list-bounces@iem.at [pd-list-bounces@iem.at] On Behalf Of Derek Holzer [derek@umatic.nl]
To: pd-list@iem.at
Subject: Re: [PD] new license for pidip and unauthorized WAS: pd-pidip into Debian
_______________________________________________Motion tracking can be done easily in GEM with a combination of
[pix-background] to create a histrogram showing only changed pixel
values, and [pix_blob] to track the center of gravity of those changed
values.
D.
On 12/7/10 6:43 PM, Jose Luis Santorcuato wrote:
> I don't quite understand why people do motion detection with squares
> like that. I use a generic motion detection that gets me the centre of
> the motion, and then when I rescale the values, it may split the
> possible results into bands or squares because of rounding (if I
> quantise them).
> Hi all!,
> what is the generic motion detector you are using Mathieu? I use pdp...
> Best Regards
> José
> 2010/12/7 Mathieu Bouchard <matju@artengine.ca <mailto:matju@artengine.ca>>:
> > On Tue, 7 Dec 2010, Pagano, Patrick wrote:
> >
> >> pdp_mgrid
> >
> > I don't quite understand why people do motion detection with squares like
> > that. I use a generic motion detection that gets me the centre of the
> > motion, and then when I rescale the values, it may split the possible
> > results into bands or squares because of rounding (if I quantise them).
> >
> > Nevertheless, if I have a picture of size (240 320 3), I can make
> groups of
> > rows and columns, for example I [#redim (15 16 20 16 3)], which turns a
> > picture into a 5-dimensional grid, and then if I [#transpose 1 2], I will
> > have a 5-dimensional grid that is a 2-dimensional grid of 16x16
> icons. After
> > that, many effects and analyses that GF can do on pictures will also
> work on
> > such groups of pictures without any change.
> >
> > But more simply, if I want an average of each 16x16 icon, I don't
> need to go
> > 5-dimensional, I can just do [#downscale_by 16 smoothly] and that's all.
> >
> >> pdp_cmap
> >> pdp_ctrack
> >> pdp_ascii
> >> are essential
> >
> > I'd like to know what it takes for you to call something «essential».
> Do you
> > mean that there is no other way to do it with Pd, or just no other
> way to do
> > with with PDP ?
> >
> >> pdp_qt will be obsolete when we get either pix_film OR pix_movie or
> >> whatever to play audio directly from a movie without extra table banging
> >> [hopefully this exists somewhere]
> >
> > There's [#in~], which is rather experimental. It uses [#to~].
> >
> > _______________________________________________________________________
> > | Mathieu Bouchard ---- tél: +1.514.383.3801 ---- Villeray, Montréal, QC
> > _______________________________________________
> > Pd-list@iem.at <mailto:Pd-list@iem.at> mailing list
> > UNSUBSCRIBE and account-management ->
> > http://lists.puredata.info/listinfo/pd-list
> >
> >
> --
> http://arselectronicachile.blogspot.com
> http://comunicacionnativa.blogspot.com/
> http://www.myspace.com/santorcuato
>
>
>
> _______________________________________________
> Pd-list@iem.at mailing list
> UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
--
::: derek holzer ::: http://macumbista.net :::
---Oblique Strategy # 207:
"Take away as much mystery as possible. What is left?"
_______________________________________________
Pd-list@iem.at mailing list
UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list
UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list