Measurement wrappers for distribution images and probability images. More...
#include <vcl_cassert.h>#include <vil/vil_image_view.h>#include <vbl/vbl_array_2d.h>#include "bbgm_image_of.h"#include "bbgm_planes_to_sample.h"Go to the source code of this file.
Functions | |
| template<class dist_ , class measure_functor_ > | |
| void | measure (bbgm_image_of< dist_ > &dimg, const vil_image_view< typename dist_::math_type > &image, vil_image_view< typename dist_::math_type > &result, const measure_functor_ &prop, typename dist_::math_type delta) |
| measure a property (typically probability) requiring a tolerance, delta. | |
| template<class dist_ , class measure_functor_ > | |
| void | measure_bkground (bbgm_image_of< dist_ > &dimg, const vil_image_view< typename dist_::math_type > &image, vil_image_view< typename dist_::math_type > &result, const measure_functor_ &prop, typename dist_::math_type pf) |
| measure the probability of background given a fixed foreground probability (pf) at each pixel. | |
| template<class dist_ , class measure_functor_ > | |
| void | measure (bbgm_image_of< dist_ > &dimg, const vil_image_view< typename dist_::math_type > &image, vil_image_view< typename dist_::math_type > &var, vil_image_view< typename dist_::math_type > &result, const measure_functor_ &prop, typename dist_::math_type delta) |
| measure a property (typically probability) requiring a tolerance, delta. | |
| template<class dist_ , class measure_functor_ > | |
| void | measure (bbgm_image_of< dist_ > &dimg, const vil_image_view< typename dist_::math_type > &image, const vbl_array_2d< typename dist_::dist_type::covar_type > &covar, vil_image_view< typename dist_::math_type > &result, const measure_functor_ &prop, const typename dist_::math_type delta) |
| measure a property (typically probability) requiring a tolerance, delta. | |
| template<class dist_ , class measure_functor_ > | |
| void | measure (bbgm_image_of< dist_ > &dimg, vil_image_view< typename dist_::math_type > &result, const measure_functor_ &prop, typename dist_::vector_type min_value, typename dist_::vector_type max_value) |
| measure a property (typically probability) requiring a tolerance. | |
Measurement wrappers for distribution images and probability images.
Modifications
<none yet>
Definition in file bbgm_measure.h.
| void measure | ( | bbgm_image_of< dist_ > & | dimg, |
| const vil_image_view< typename dist_::math_type > & | image, | ||
| vil_image_view< typename dist_::math_type > & | result, | ||
| const measure_functor_ & | prop, | ||
| typename dist_::math_type | delta | ||
| ) |
measure a property (typically probability) requiring a tolerance, delta.
For probability, delta defines a square n-dimensional box over which the probability density is integrated. For example, the measure_functor_ is the bsta_probability_functor.
Definition at line 26 of file bbgm_measure.h.
| void measure | ( | bbgm_image_of< dist_ > & | dimg, |
| const vil_image_view< typename dist_::math_type > & | image, | ||
| vil_image_view< typename dist_::math_type > & | var, | ||
| vil_image_view< typename dist_::math_type > & | result, | ||
| const measure_functor_ & | prop, | ||
| typename dist_::math_type | delta | ||
| ) |
measure a property (typically probability) requiring a tolerance, delta.
For probability, delta defines a square n-dimensional box over which the probability density is integrated. For example, the measure_functor_ is the bsta_probability_functor. In this method the tolerance is augmented by an additional value that varies with position in the image
Definition at line 112 of file bbgm_measure.h.
| void measure | ( | bbgm_image_of< dist_ > & | dimg, |
| const vil_image_view< typename dist_::math_type > & | image, | ||
| const vbl_array_2d< typename dist_::dist_type::covar_type > & | covar, | ||
| vil_image_view< typename dist_::math_type > & | result, | ||
| const measure_functor_ & | prop, | ||
| const typename dist_::math_type | delta | ||
| ) |
measure a property (typically probability) requiring a tolerance, delta.
For probability, delta defines a square n-dimensional box over which the probability density is integrated. As an example, the measure_functor_ is the bsta_probability_functor. In this method, the covariance of the distribution is augmented by and additional amount
Definition at line 156 of file bbgm_measure.h.
| void measure | ( | bbgm_image_of< dist_ > & | dimg, |
| vil_image_view< typename dist_::math_type > & | result, | ||
| const measure_functor_ & | prop, | ||
| typename dist_::vector_type | min_value, | ||
| typename dist_::vector_type | max_value | ||
| ) |
measure a property (typically probability) requiring a tolerance.
The tolerance is defined by a minimum and maximum values For probability, min and max defines a square n-dimensional box over which the probability density is integrated. This functor assumes that the same box applies to the entire distribution image. An example is probability of area lying within an interval, where the distribution image is a 2-d array of area probability densities.
Definition at line 202 of file bbgm_measure.h.
| void measure_bkground | ( | bbgm_image_of< dist_ > & | dimg, |
| const vil_image_view< typename dist_::math_type > & | image, | ||
| vil_image_view< typename dist_::math_type > & | result, | ||
| const measure_functor_ & | prop, | ||
| typename dist_::math_type | pf | ||
| ) |
measure the probability of background given a fixed foreground probability (pf) at each pixel.
Bayes rule gives
p(I|b)P(b) p(I|b)(1-pf) p(I|b)(1-pf) P(b) = ---------------------- = -------------------- = ----------------- p(I|b)P(b) + p(I|f)P(f) p(I|b)(1-pf)+ p(I|f)pf p(I|b)(1-pf)+ pf
the foreground probability is assumed to be uniform on the interval (0,1)
Definition at line 71 of file bbgm_measure.h.
1.7.5.1