Re: [Meshlab-devel] detail transfer plugin

[skunkwerk <sku...@gm...>]

it seems as if both of the parametrization algorithms i've seen
(least-squares conformal mapping and angle-based flattening) require meshes
with a genus > 0 to be cut (so as to be topologically equivalent to a
disc).  i've read some papers on methods of doing this automatically, but
haven't seen many implementations out there.  is there a method in meshlab
to traverse over edges and create such cuts?  any suggestions on simple,
non-optimal algorithms to use as a starting point?

thanks,
imran

On Sun, Aug 24, 2008 at 9:13 AM, skunkwerk <sku...@gm...> wrote:

> haha,
>    i'll be the first to admit i'm not the greatest programmer on the
> planet.  but i learn fast, and am determined enough to spend however much
> time it takes.  it couldn't hurt if i try to work on this, could it?
> perhaps you could assign one of your students in the new term to take the
> lead?
>
> it seems like the CGAL library has a clean interface to LSCM - you input a
> mesh, and it outputs a uv pair for each vertex:
>
> http://www.cgal.org/Manual/3.3/doc_html/cgal_manual/Surface_mesh_parameterization/Chapter_main.html
>
> imran
>
> On Sat, Aug 23, 2008 at 5:06 PM, Paolo Cignoni <pao...@is...>wrote:
>
>>  Dear Imran,
>>
>> sorry for speaking so frankly, but from many of your last emails I got the
>> precise feeling that detail transfer and unwrapping are a task too hard for
>> you.
>> Your last email revealed that probably you still have C++ doubts at a
>> basic level.
>> Particularly I am referring to your last email (but others examples could
>> be found) where you said:
>> ...
>> skunkwerk wrote:
>> > I've noticed when you're random-accessing vertices, you apparently need
>> to do: m.vert[f.V(c)-&(*m.vert.begin())], where f is the face and V is the
>> vertex.
>> > 1)  I'm not quite sure while you need the m.vert.begin, as doesn't f.V()
>> hold the absolute index of the vertex you want?
>> > 2)  And if you do need it, wouldn't you also need to do a similar thing
>> for random-accessing faces too?
>> > Because all I'm doing right now is: m.face[(*fit)], where fit is a face
>> iterator.
>> ...
>>
>> Teaching a CG class at the university since 1997 I think that I can say
>> you that you should invest a bit more time on studying the C++ basis
>> following a structured course book instead of jumping directly in the middle
>> of a rather complex C++ system based on a convoluted templated library. That
>> would surely shorten the time you spend in coding and debugging in C++.
>>
>> If you really want to contribute, spend a bit more time in testing
>> debugging the whole meshlab, not only your plugin, e.g. try to find crash
>> points, and writing down repeatable sequence of actions that lead in a
>> deterministic way to a crash. and leave to more experienced coders the
>> burden of patch it in the best way (if possible).
>> That would be really a valuable contribution to the whole project.
>>
>> Thanks
>>
>> p.
>>
>>
>>
>>
>> skunkwerk wrote:
>>
>> Paolo,
>>    i'd like to again request your permission to start work on the normal
>> mapping/detail transfer plugin.  i can continue to try and fix the last
>> remaining bug with the texture atlas (when you increase its size), but I
>> don't see any way to make it faster besides removing the qDebug statements
>> from it... and I don't have anything else left to work on anyway :0
>>
>> would the first place to start be the filter_sampling.cpp file, to add
>> normals as an option to the attribute transfer?
>> i've been looking at the lscm code to unwrap models... Blender seems to be
>> using the superLU library as a numerical solver.  I'm not sure how to
>> interface with it the code at this point...
>>
>> thanks,
>> imran
>>
>> 2008/7/22 Paolo Cignoni <pao...@is...>
>>
>>> No.
>>>
>>> You have to choose one of the following possibility
>>> - select more than a single file in the open dialog
>>> - open a mesh and then use the open as a new layer menu entry
>>> - open a mesh and press the "+" button.
>>>
>>> http://meshlab.sourceforge.net/wiki/index.php/Layers
>>>
>>>
>>> skunkwerk wrote:
>>>
>>>  ah.. cool.  I opened up the same model twice (so they're 2 windows now) but
>>> still don't see more than one mesh option in the hausdorff distance
>>> filter... am i doing something wrong?
>>>
>>> On Mon, Jul 21, 2008 at 10:46 PM, Paolo Cignoni <pao...@is...> <pao...@is...>
>>> wrote:
>>>
>>>
>>>
>>>   skunkwerk wrote:
>>>
>>> hey paolo,
>>>      i was looking at the filter_sampling plugins - both the distance and
>>> 'vertex attribute transfer' plugins require 2 meshes to compare - how would
>>> you get meshlab to load 2 models, such as one of the original and one of the
>>> model after a filter has run on it?
>>>
>>>
>>>  Dont tell me that you have missed the whole layer mechanism...
>>> it was one of the main feature of the 1.1 version...
>>> http://meshlab.sourceforge.net/wiki/index.php/Release_Notes_1.1.0
>>>
>>>
>>>  would normal mapping require adding a "normals" option to the vertex
>>> attribute transfer plugin?  then you'd have to bake the normals into a
>>> texture, right?
>>>
>>>
>>>
>>>  in a short future,
>>> yes.
>>>
>>>  imran
>>>
>>> 2008/6/25 Paolo Cignoni <pao...@is...> <pao...@is...> <pao...@is...> <pao...@is...>:
>>>
>>>
>>>
>>>   skunkwerk wrote:
>>>
>>> hi Paolo,
>>>     i was looking at software that could do normal mapping and noticed it
>>> was listed as one of the plugin projects in meshlab.
>>>
>>> is anyone working on it yet?  anyone else interested in helping?
>>>
>>> I read your paper on normal mapping from a few years ago.  Do I need to get
>>> the texture atlas plugin fixed before starting on the detail transfer?
>>>
>>>  yes complete the texture atlas first.
>>>
>>> about the redetailing stuff: it is almost in. If you look at the new
>>> filter_sampling tool you can find all the basic pieces: closest retrieval
>>> and texel sampling.
>>> I think that adding a attribute transfer is matter of a few lines of code.
>>> consider it already done.
>>> (yes i have to remove it from the projecgts :) )
>>> p.
>>>
>>>
>>>
>>> imran
>>>
>>> ------------------------------
>>>
>>> -------------------------------------------------------------------------
>>> Check out the new SourceForge.net Marketplace.
>>> It's the best place to buy or sell services for
>>> just about anything Open Source.http://sourceforge.net/services/buy/index.php
>>>
>>> ------------------------------
>>>
>>> ______________________________
>>> _________________
>>> Meshlab-devel mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/meshlab-devel
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> --
>>> Paolo Cignoni  -- Senior Researcher
>>> Visual Computing Laboratory - ISTI - CNR http://vcg.isti.cnr.it/~cignoni <http://vcg.isti.cnr.it/%7Ecignoni> <http://vcg.isti.cnr.it/%7Ecignoni> <http://vcg.isti.cnr.it/%7Ecignoni> <http://vcg.isti.cnr.it/%7Ecignoni> <http://vcg.isti.cnr.it/%7Ecignoni> <http://vcg.isti.cnr.it/%7Ecignoni> <http://vcg.isti.cnr.it/%7Ecignoni>
>>>
>>>
>>>
>>>
>>> ISTI - CNR
>>> Via Moruzzi 1,
>>> 56124 Pisa
>>> ITALY
>>>
>>>
>>> -------------------------------------------------------------------------
>>> Check out the new SourceForge.net Marketplace.
>>> It's the best place to buy or sell services for
>>> just about anything Open Source.http://sourceforge.net/services/buy/index.php
>>> ______________________________
>>> _________________
>>> Meshlab-devel mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/meshlab-devel
>>>
>>>
>>>        ------------------------------
>>>
>>> -------------------------------------------------------------------------
>>> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
>>> Build the coolest Linux based applications with Moblin SDK & win great prizes
>>> Grand prize is a trip for two to an Open Source event anywhere in the worldhttp://moblin-contest.org/redirect.php?banner_id=100&url=/
>>>
>>> ------------------------------
>>>
>>> _______________________________________________
>>> Meshlab-devel mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/meshlab-devel
>>>
>>>
>>>
>>>
>>>
>>>
>>> --
>>> Paolo Cignoni  -- Senior Researcher
>>> Visual Computing Laboratory - ISTI - CNR http://vcg.isti.cnr.it/~cignoni <http://vcg.isti.cnr.it/%7Ecignoni> <http://vcg.isti.cnr.it/%7Ecignoni> <http://vcg.isti.cnr.it/%7Ecignoni>
>>>
>>> ISTI - CNR
>>> Via Moruzzi 1,
>>> 56124 Pisa
>>> ITALY
>>>
>>>
>>> ------------------------------------------------------------
>>> -------------
>>> This SF.Net email is sponsored by the Moblin Your Move Developer's
>>> challenge
>>> Build the coolest Linux based applications with Moblin SDK & win great
>>> prizes
>>> Grand prize is a trip for two to an Open Source event anywhere in the worldhttp://moblin-contest.org/redirect.php?banner_id=100&url=/
>>> _______________________________________________
>>> Meshlab-devel mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/meshlab-devel
>>>
>>>
>>>        ------------------------------
>>>
>>> -------------------------------------------------------------------------
>>> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
>>> Build the coolest Linux based applications with Moblin SDK & win great prizes
>>> Grand prize is a trip for two to an Open Source event anywhere in the worldhttp://moblin-contest.org/redirect.php?banner_id=100&url=/
>>>
>>> ------------------------------
>>>
>>> _______________________________________________
>>> Meshlab-devel mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/meshlab-devel
>>>
>>>
>>>
>>> --
>>> Paolo Cignoni  -- Senior Researcher
>>> Visual Computing Laboratory - ISTI - CNR http://vcg.isti.cnr.it/~cignoni <http://vcg.isti.cnr.it/%7Ecignoni>
>>>
>>> ISTI - CNR
>>> Via Moruzzi 1,
>>> 56124 Pisa
>>> ITALY
>>>
>>>
>>> -------------------------------------------------------------------------
>>> This SF.Net email is sponsored by the Moblin Your Move Developer's
>>> challenge
>>> Build the coolest Linux based applications with Moblin SDK & win great
>>> prizes
>>> Grand prize is a trip for two to an Open Source event anywhere in the
>>> world
>>> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>>> _______________________________________________
>>> Meshlab-devel mailing list
>>> Mes...@li...
>>> https://lists.sourceforge.net/lists/listinfo/meshlab-devel
>>>
>>>
>>  ------------------------------
>>
>> -------------------------------------------------------------------------
>> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
>> Build the coolest Linux based applications with Moblin SDK & win great prizes
>> Grand prize is a trip for two to an Open Source event anywhere in the worldhttp://moblin-contest.org/redirect.php?banner_id=100&url=/
>>
>> ------------------------------
>>
>> _______________________________________________
>> Meshlab-devel mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/meshlab-devel
>>
>>
>>
>> --
>> Paolo Cignoni  -- Senior Researcher
>> Visual Computing Laboratory - ISTI - CNR http://vcg.isti.cnr.it/~cignoni <http://vcg.isti.cnr.it/%7Ecignoni>
>>
>> ISTI - CNR
>> Via Moruzzi 1,
>> 56124 Pisa
>> ITALY
>>
>>
>> -------------------------------------------------------------------------
>> This SF.Net email is sponsored by the Moblin Your Move Developer's
>> challenge
>> Build the coolest Linux based applications with Moblin SDK & win great
>> prizes
>> Grand prize is a trip for two to an Open Source event anywhere in the
>> world
>> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>> _______________________________________________
>> Meshlab-devel mailing list
>> Mes...@li...
>> https://lists.sourceforge.net/lists/listinfo/meshlab-devel
>>
>>
>