Point Cloud Library (PCL)  1.14.1
List of all members | Public Member Functions | Public Attributes
pcl::NarfKeypoint::Parameters Struct Reference

Parameters used in this class. More...

#include <pcl/keypoints/narf_keypoint.h>

Public Member Functions

 Parameters ()=default
 

Public Attributes

float support_size {-1.0f}
 This defines the area 'covered' by an interest point (in meters) More...
 
int max_no_of_interest_points {-1}
 The maximum number of interest points that will be returned. More...
 
float min_distance_between_interest_points {0.25f}
 Minimum distance between maximas (this is a factor for support_size, i.e. More...
 
float optimal_distance_to_high_surface_change {0.25}
 The distance we want keep between keypoints and areas of high surface change (this is a factor for support_size, i.e., the distance is optimal_distance_to_high_surface_change*support_size) More...
 
float min_interest_value {0.45f}
 The minimum value to consider a point as an interest point. More...
 
float min_surface_change_score {0.2f}
 The minimum value of the surface change score to consider a point. More...
 
int optimal_range_image_patch_size {10}
 The size (in pixels) of the image patches from which the interest value should be computed. More...
 
float distance_for_additional_points {0.0f}
 All points in this distance to a found maximum, that are above min_interest_value are also added as interest points (this is a factor for support_size, i.e. More...
 
bool add_points_on_straight_edges {false}
 If this is set to true, there will also be interest points on straight edges, e.g., just indicating an area of high surface change. More...
 
bool do_non_maximum_suppression {true}
 If this is set to false there will be much more points (can be used to spread points over the whole scene (combined with a low min_interest_value)) More...
 
bool no_of_polynomial_approximations_per_point {false}
 If this is >0, the exact position of the interest point is determined using bivariate polynomial approximations of the interest values of the area. More...
 
int max_no_of_threads {1}
 The maximum number of threads this code is allowed to use with OPNEMP. More...
 
bool use_recursive_scale_reduction {false}
 Try to decrease runtime by extracting interest points at lower reolution in areas that contain enough points, i.e., have lower range. More...
 
bool calculate_sparse_interest_image {true}
 Use some heuristics to decide which areas of the interest image can be left out to improve the runtime. More...
 

Detailed Description

Parameters used in this class.

Definition at line 73 of file narf_keypoint.h.

Constructor & Destructor Documentation

pcl::NarfKeypoint::Parameters::Parameters ( )
default

Member Data Documentation

bool pcl::NarfKeypoint::Parameters::add_points_on_straight_edges {false}

If this is set to true, there will also be interest points on straight edges, e.g., just indicating an area of high surface change.

Definition at line 97 of file narf_keypoint.h.

bool pcl::NarfKeypoint::Parameters::calculate_sparse_interest_image {true}

Use some heuristics to decide which areas of the interest image can be left out to improve the runtime.

Definition at line 108 of file narf_keypoint.h.

float pcl::NarfKeypoint::Parameters::distance_for_additional_points {0.0f}

All points in this distance to a found maximum, that are above min_interest_value are also added as interest points (this is a factor for support_size, i.e.

the distance is distance_for_additional_points*support_size)

Definition at line 93 of file narf_keypoint.h.

Referenced by pcl::operator<<().

bool pcl::NarfKeypoint::Parameters::do_non_maximum_suppression {true}

If this is set to false there will be much more points (can be used to spread points over the whole scene (combined with a low min_interest_value))

Definition at line 99 of file narf_keypoint.h.

int pcl::NarfKeypoint::Parameters::max_no_of_interest_points {-1}

The maximum number of interest points that will be returned.

Definition at line 78 of file narf_keypoint.h.

int pcl::NarfKeypoint::Parameters::max_no_of_threads {1}

The maximum number of threads this code is allowed to use with OPNEMP.

Definition at line 105 of file narf_keypoint.h.

float pcl::NarfKeypoint::Parameters::min_distance_between_interest_points {0.25f}

Minimum distance between maximas (this is a factor for support_size, i.e.

the distance is min_distance_between_interest_points*support_size)

Definition at line 79 of file narf_keypoint.h.

Referenced by pcl::operator<<().

float pcl::NarfKeypoint::Parameters::min_interest_value {0.45f}

The minimum value to consider a point as an interest point.

Definition at line 86 of file narf_keypoint.h.

Referenced by pcl::operator<<().

float pcl::NarfKeypoint::Parameters::min_surface_change_score {0.2f}

The minimum value of the surface change score to consider a point.

Definition at line 87 of file narf_keypoint.h.

bool pcl::NarfKeypoint::Parameters::no_of_polynomial_approximations_per_point {false}

If this is >0, the exact position of the interest point is determined using bivariate polynomial approximations of the interest values of the area.

Definition at line 102 of file narf_keypoint.h.

float pcl::NarfKeypoint::Parameters::optimal_distance_to_high_surface_change {0.25}

The distance we want keep between keypoints and areas of high surface change (this is a factor for support_size, i.e., the distance is optimal_distance_to_high_surface_change*support_size)

Definition at line 82 of file narf_keypoint.h.

int pcl::NarfKeypoint::Parameters::optimal_range_image_patch_size {10}

The size (in pixels) of the image patches from which the interest value should be computed.

This influences, which range image is selected from the scale space to compute the interest value of a pixel at a certain distance.

Definition at line 88 of file narf_keypoint.h.

float pcl::NarfKeypoint::Parameters::support_size {-1.0f}

This defines the area 'covered' by an interest point (in meters)

Definition at line 77 of file narf_keypoint.h.

Referenced by pcl::operator<<().

bool pcl::NarfKeypoint::Parameters::use_recursive_scale_reduction {false}

Try to decrease runtime by extracting interest points at lower reolution in areas that contain enough points, i.e., have lower range.

Definition at line 106 of file narf_keypoint.h.


The documentation for this struct was generated from the following file: