Obviously this will be limited to "full" objects I presume.
I found a "2D to 3D image converter" but hey it's not top top eh ...
Thank you friends!
with several photos taken under different angles you can locate the point clearly visible in space because there are sharp angles: for curved surfaces and uniform colors there is no solution with several photos
solution, projector, which projects a grid on the object to be photographed, and the geometric solution becomes obvious, and it becomes possible to make a software
the projector can be a slide projector, with a grid slide, or a video projector to more easily change the fineness of the grid
while waiting for a special software we can do the geometric construction with solidworks to place in space each angle of the grid visible on a photo
ggdorm wrote:following a quick test, the result is not very famous.
Thank you for the feedback, can you tell us more by "not very famous"?
Can you give us images of the object and the modeling?
I unfortunately have no more photos. I had made a brief test a short year ago and have since changed my PC. I discovered this software via a site which presents the construction of a "3D scanner" here: http://www.instructables.com/id/123D-Sc ... ner/#intro
The problem I had was that there are areas where the texture was not restored but I had not proceeded as explained on this site. By repeating the test more conscientiously, the results will surely be better.
I found this on my side: VI3DIM 3D RECONSTRUCTION
I will test the free version: http://www.vi3dim.com/ but it's not free at all ($ 400)
by taking around thirty photos by rotating 360 ° around the object, we can obtain a stl file or a point cloud.
I did a few tests and need to work on the lighting, the background textures, ...
otherwise, it's free and it runs on an online platform. So no installation to do but it takes a few hours of treatment ...
the conversion of a set of photos into a 3D model is based on the geometric principles of photogrammetry, there are several free software including:
- -the couple OSMBundler / Python Photogrammetry Toolbox which has the advantage of being simple to use and having a graphical interface http://www.arc-team.homelinux.com/arcteam/ppt.php
- MICMAC which is a software developed by IGN, it requires much more handling but allows quality results http://www.micmac.ign.fr/svn/micmac_dat ... Interface/
- OpenMVG, which does not have a graphical interface (therefore only in command line) but which is easy to use and which gives good results https://github.com/openMVG/openMVG
These software all produce point clouds (PLY file) which can be edited by MeshLab ( http://meshlab.sourceforge.net/ ) the conversion into a mesh object is necessary for the transition to printing, it can be done either in MeshLab or with a more efficient utility: FishRecon ( http://www.cs.jhu.edu/~misha/Code/PoissonRecon ).
And finally, an IRC (English-speaking, Italian-speaking and French-speaking) discussion channel where this subject is discussed: #archeos on irc.freenode.net
Who is online ?
Users browsing this forum : No registered users and 1 guest