A class for representing a set of sites in an MRF. More...
#include <sdet/sdet_mrf_site_bp_sptr.h>#include <vil/vil_image_resource.h>#include <vil/vil_image_view.h>#include <vbl/vbl_array_2d.h>#include <vbl/vbl_ref_count.h>#include <vcl_vector.h>#include <sdet/sdet_mrf_bp_sptr.h>Go to the source code of this file.
Classes | |
| class | sdet_mrf_bp |
Functions | |
| void | lower_envelope_linear (float w, vcl_vector< float > &msg) |
| computes the lower envelope of a message array with $V(fp, fq) = w|fp-fq|$. | |
| vcl_vector< float > | lower_envelope_quadratic (float w, vcl_vector< float > const &h) |
| computes the lower envelope of a message array with $V(fp, fq) = w(fp-fq)^2$. | |
A class for representing a set of sites in an MRF.
Each MRF site has a 4-neighborhood, u l x r d with index (u, l, r, d) ==> (0, 1, 2, 3)
Definition in file sdet_mrf_bp.h.
| void lower_envelope_linear | ( | float | w, |
| vcl_vector< float > & | msg | ||
| ) |
computes the lower envelope of a message array with $V(fp, fq) = w|fp-fq|$.
Computes in place
Definition at line 17 of file sdet_mrf_bp.cxx.
| vcl_vector<float> lower_envelope_quadratic | ( | float | w, |
| vcl_vector< float > const & | h | ||
| ) |
computes the lower envelope of a message array with $V(fp, fq) = w(fp-fq)^2$.
Used in current implementation
Definition at line 40 of file sdet_mrf_bp.cxx.
1.7.5.1