It seems as though the artengine cvs and the gridflow page is down. Where can I get the latest gridflow incliding your patch called #labeling? I would like to get this going. Does your #labeling patch output blob coordinates in a matrix style? Thanks for your help, I'm gonna try hard to make gridflow work os x. Alain
From: Mathieu Bouchard matju@artengine.ca Date: 2007/05/17 Thu PM 02:38:32 EDT To: nosehair911@bellsouth.net Subject: Re: [GEM-dev] best way to wrap opencv library for noob
On Thu, 17 May 2007, nosehair911@bellsouth.net wrote:
I am trying to get an FTIR setup going and I need a good motion tracker for os 10.4.9. I cant get gridflow to compile,
Then subscribe and write to gridflow-dev@artengine.ca (http://lists.artengine.ca/mailman/listinfo/gridflow-dev)
Or write to me
Or use online chatting: IRC chat.freenode.net #dataflow
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal QC Canada
On Thu, 17 May 2007, nosehair911@bellsouth.net wrote:
It seems as though the artengine cvs and the gridflow page is down.
Ok, I just fixed the gridflow anonymous cvs, which has been down since monday evening, because I had to reinstall artengine.ca and forgot to reinstall it. (However, this disabled the non-anonymous cvs and I have to figure out why)
The GridFlow page was down from Monday to Wednesday for other related reasons, but if you have tried today it should have worked. Else could you please send me more info about it?
Where can I get the latest gridflow incliding your patch called #labeling?
It should be working right now. from extra/, do: mkdir gridflow cd gridflow export CVSROOT=:pserver:anonymous@cvs.gridflow.ca:/home/cvs/gridflow cvs login cvs checkout .
I would like to get this going. Does your #labeling patch output blob coordinates in a matrix style?
For each blob it outputs one scalar for the number of pixels in the blob, a vector for the centre of the blob, and a matrix for the variance of the blob. The variance can be turned into a standard deviation made either of xx, xy, yy components (cartesian), or long radius, short radius, angle (polar), if you use appropriate math objects. After sending this for every blob, it outputs one blob in index colour where each blob has a different colour and the background is 0. The input must be a binary image made of only 0 and 1.
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal QC Canada
On Thu, 17 May 2007, nosehair911@bellsouth.net wrote:
Thanks for your help, I'm gonna try hard to make gridflow work os x.
Oh. This is OSX ? GridFlow is slower than normal on OSX G3/G4/G5, but I don't know whether this problem extends to OSX intel.
Also, the GEM/GridFlow conversion objects don't work because most versions of Pd on OSX prevent externals from referring to each other's contents, because GridFlow needs to assign to variables CPPExtern::m_holder and CPPExtern::m_holdname.
If a GEM expert may help me: it needs to do so in this way:
#ifdef HAVE_GEM CPPExtern::m_holder = (t_object *)bself; #ifdef HAVE_HOLDNAME CPPExtern::m_holdname = "keep_gem_happy"; #endif ... #ifdef HAVE_GEM ... CPPExtern::m_holder = NULL; #ifdef HAVE_HOLDNAME CPPExtern::m_holdname=NULL; #endif
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal QC Canada
-
Mathieu Bouchard -
nosehair911＠bellsouth.net