Optimized Belief Propagation (CPU and GPU)
DisparityMapEvaluation Class Reference

Class to store disparity map evaluation results. Specifically comparison between two disparity maps such as output disparity map from bp processing and ground truth disparity map. More...

#include <DisparityMapEvaluation.h>

Public Member Functions

void InitializeWithEvalParams (const beliefprop::DisparityMapEvaluationParams &eval_params)
 Initialize evaluation results with evaluation parameters. More...
 
RunData AsRunData () const
 Retrieve evaluation results as RunData. More...
 

Public Attributes

std::array< float, 2 > average_disp_abs_diff_no_max_w_max_ {0, 0}
 Total and average value of the absolute difference between the disparity values for all pixels in disparity images 1 and 2 with and without maximum disparity difference at each pixel. More...
 
float disparity_error_max_ {std::numeric_limits<float>::max()}
 
std::map< float, float > prop_sig_diff_pixels_at_thresholds_
 Proportion of pixels where the difference between the disparity values in disparity images 1 and 2 is greater than SIG_DIFF_THRESHOLD_STEREO_EVAL (not including border regions) More...
 
std::map< float, unsigned int > num_sig_diff_pixels_at_thresholds_
 Stores the number of pixels where the difference between the disparity values in disparity images 1 and 2 is greater than SIG_DIFF_THRESHOLD_STEREO_EVAL. More...
 

Detailed Description

Class to store disparity map evaluation results. Specifically comparison between two disparity maps such as output disparity map from bp processing and ground truth disparity map.

Definition at line 74 of file DisparityMapEvaluation.h.

Member Function Documentation

◆ AsRunData()

RunData DisparityMapEvaluation::AsRunData ( ) const

Retrieve evaluation results as RunData.

Returns
RunData object containing evaluation results

Definition at line 42 of file DisparityMapEvaluation.cpp.

Here is the call graph for this function:

◆ InitializeWithEvalParams()

void DisparityMapEvaluation::InitializeWithEvalParams ( const beliefprop::DisparityMapEvaluationParams eval_params)

Initialize evaluation results with evaluation parameters.

Parameters
eval_params

Definition at line 30 of file DisparityMapEvaluation.cpp.

Here is the caller graph for this function:

Member Data Documentation

◆ average_disp_abs_diff_no_max_w_max_

std::array<float, 2> DisparityMapEvaluation::average_disp_abs_diff_no_max_w_max_ {0, 0}

Total and average value of the absolute difference between the disparity values for all pixels in disparity images 1 and 2 with and without maximum disparity difference at each pixel.

Definition at line 94 of file DisparityMapEvaluation.h.

◆ disparity_error_max_

float DisparityMapEvaluation::disparity_error_max_ {std::numeric_limits<float>::max()}

Definition at line 95 of file DisparityMapEvaluation.h.

◆ num_sig_diff_pixels_at_thresholds_

std::map<float, unsigned int> DisparityMapEvaluation::num_sig_diff_pixels_at_thresholds_

Stores the number of pixels where the difference between the disparity values in disparity images 1 and 2 is greater than SIG_DIFF_THRESHOLD_STEREO_EVAL.

Definition at line 105 of file DisparityMapEvaluation.h.

◆ prop_sig_diff_pixels_at_thresholds_

std::map<float, float> DisparityMapEvaluation::prop_sig_diff_pixels_at_thresholds_

Proportion of pixels where the difference between the disparity values in disparity images 1 and 2 is greater than SIG_DIFF_THRESHOLD_STEREO_EVAL (not including border regions)

Definition at line 100 of file DisparityMapEvaluation.h.


The documentation for this class was generated from the following files: