Ambrosio-Tortorelli package #280
Conversation
|
Looks like the problem with Appveyor is related to boost program options. Perhaps there is something specific to do for VC++ ? |
|
Strange.. appveyor was fine with boost_programpotions in previous PR.. @kerautret, would you have an idea? |
|
I look it but yes that is strange! ;( |
| @@ -47,7 +47,7 @@ SET(DGTALTOOLS_VERSION "${DGtalTools_VERSION_MAJOR}.${DGtalTools_VERSION_MINOR}. | |||
| SET(PROJECT_VERSION "${DGtalTools_VERSION_MAJOR}.${DGtalTools_VERSION_MINOR}.${DGtalTools_VERSION_PATCH}") | |||
| SET(VERSION ${DGtalTools_VERSION_MAJOR}.${DGtalTools_VERSION_MINOR}.${DGtalTools_VERSION_PATCH}) | |||
|
|
|||
| SET(Boost_USE_STATIC_LIBS ON) | |||
| SET(Boost_USE_STATIC_LIBS OFF) | |||
| SET(Boost_USE_MULTITHREADED ON) | |||
There was a problem hiding this comment.
@JacquesOlivierLachaud I wonder if it is not related to this change (boost recommend to set it to true)
There was a problem hiding this comment.
It may be that. I have to change it because I have a problem with boost static lib on my laptop.
I will try that.
|
the new error looks to be related to eigen that I have not tested in appveyor. |
|
Yes. I forgot to specify that we need Eigen for these two tools (because DEC needs Eigen). |
| Discrete Hodge star operators \f$ \mathbf{\star} \f$ send \f$ k \f$-forms | ||
| of the primal complex \a K onto \f$ (n-k) \f$-forms of the dual complex | ||
| \a K'. | ||
|
|
There was a problem hiding this comment.
perhaps precise also: $\bar{ \mathbf{\star} \f} $ for the dual .
|
|
||
| A second possibility is to define \f$ u \f$ and \f$ g \f$ on the vertices | ||
| and \f$ v \f$ on the edges. We denote this formulation | ||
| AT10. Gray levels are seen as point mass on the center of pixels, so |
| is a dual 1-form in between \f$ u \f$. It follows: | ||
| \f[ | ||
| \displaystyle | ||
| AT10(u,v) = \Sigma_{i=1}^n |
| & (2.2) | ||
| \end{array}\right. | ||
| \f] | ||
| for the derivative of AT10. Since all matrices are |
| OPTIMIZE-AT01( (g_1, ..., g_N ): 0-forms, | ||
| (\alpha, \lambda, \varepsilon_1, \varepsilon_2, \varepsilon_r): reals ); | ||
| \f$ | ||
|
|
There was a problem hiding this comment.
Here after it proceed that on my doc: (with doxyen 1.8.11)
|
|
||
| @subsection AT3_1 Some examples of restoration and discontinuity detections | ||
|
|
||
| We display below the different resconstruction of image "lena" with |
|
|
||
| [3] Marion Foare, Jacques-Olivier Lachaud, and Hugues Talbot. "Numerical | ||
| implementation of the Ambrosio-Tortorelli functional using discrete | ||
| calculus and application to image restoration and inpainting." ???? |
There was a problem hiding this comment.
You can complete since accepted ;)
To appears to the first Workshop on Reproducible Reaseach In Pattern Recognition (RRPR 2016), Springer LNCS
There was a problem hiding this comment.
Thanks for the review. I look at it this afternoon.
|
Every comment has been addressed. |
|
looks fine, thanks, going to merge just waiting for travis ;) |
Thanks a lot for contributing to DGtalTools, before submitting your PR, please fill up the description and make sure that all checkboxes are checked.
PR Description
This PR introduces the imageProcessing package for DGtalTools. It is composed of two programs and some helper classes to do image restoration (and inpainting) with Ambrosio-Tortorelli functional.
Checklist
Note: I propose to add a subdirectory imageProcessing
cmakemode (otherwise, Travis C.I. will fail).I think it is okay now. For the readme, I think I have to wait for a first generation of the documentation before adding snapshots.