Regularization #306
Regularization #306
Conversation
|
beside #307, this code can be reviewed |
|
nice, thanks, I will have a look once I found a moment ;) |
|
I will have a look too. For the obvious reasons ;-) |
|
;) 👌 |
| @@ -104,42 +104,40 @@ void CCCounter(Rank& r, Parent& p, const Image& elements, const unsigned int con | |||
| typename Image::Point decz(0,0,1); | |||
|
|
|||
| //Merging process | |||
| for(typename Image::Domain::ConstIterator e = elements.domain().begin(); | |||
| e !=elements.domain().end(); ++e) | |||
| for ( typename Image::Domain::ConstIterator e = elements.domain().begin(); | |||
There was a problem hiding this comment.
not sure.. looks like formatting only
| Options options; | ||
| po::options_description po_shape("File options"); | ||
| po_shape.add_options() | ||
| ("image-filename,i", po::value<std::string>(&options.image_filename)->default_value(""), "input vol filename for image shape or input cvs filename for surfels and normals") |
There was a problem hiding this comment.
For the case of vol file, what is the threshold used to consider the object ? (128?) perhaps precise it and ideally an option to change it could be nice ;)
There was a problem hiding this comment.
Have you an example cvs file example to test ?
There was a problem hiding this comment.
for the vol, the only assumption is that 0 is the background.
| po_options.add(po_shape).add(po_normal).add(po_approx).add_options() | ||
| ("help,h", "display this message") | ||
| ; | ||
|
|
There was a problem hiding this comment.
by using a vol generated from a mesh2vol, I obtained such a result:
I suppose that it is due to the fact that the volume is not filled so the thin parts overlap?
|
fine, I think that we can merge, It works fine on set of full voxels: |
|
not tested on imported normals since the conversion with surfers need a little time to check in my tools. For the thin structure I put an issue. |
PR Description
Digital Surface Regularization (DGCI2017)
Previously #299
Checklist
cmakemode (otherwise, Travis C.I. will fail).