rso::CStereoOdometryEstimator::TDetectParams Struct Reference

#include <libstereo-odometry.h>

Public Types
enum	NMSMethod { nmsmStandard, nmsmAdaptive }
enum	TDMethod { dmORB, dmFAST_ORB, dmFASTER, dmKLT }

Public Member Functions
	TDetectParams ()
void	dumpToConsole ()

Public Attributes
TDMethod	detect_method
	Method to detect features (also affects to the matching process) More...
double	target_feats_per_pixel
	Desired number of features per square pixel (Default: 10/1000) More...
int	KLT_win
	Window for the KLT response evaluation (Default: 4) More...
double	minimum_KLT_response
	Minimum KLT response to not to discard a feature as being in a textureless zone (Default: 10) More...
bool	non_maximal_suppression
	Enable/disable the non-maximal suppression after the detection (5x5 windows is used) More...
NMSMethod	nmsMethod
	Method to perform non maximal suppression. More...
size_t	min_distance
	The allowed minimun distance between features. More...
size_t	orb_nfeats
	Number of features to be detected (only for ORB) More...
size_t	orb_nlevels
	The number of pyramid levels. More...
double	minimum_ORB_response
	Minimum ORB response [Harris response] to not to discard a feature as being in a textureless zone (Default: 0.005) More...
int	fast_min_th
int	fast_max_th
	Limits for the FAST (within ORB) detector (dynamic) threshold. More...
int	initial_FAST_threshold
	The initial threshold of the FAST feature detector, which will be dynamically adapted to fulfill target_feats_per_pixel (Default=15) More...

Detailed Description

Parameters for the keypoint detection

Member Enumeration Documentation

enum rso::CStereoOdometryEstimator::TDetectParams::NMSMethod

Enumerator
nmsmStandard
nmsmAdaptive

enum rso::CStereoOdometryEstimator::TDetectParams::TDMethod

Enumerator
dmORB
dmFAST_ORB
dmFASTER
dmKLT

Constructor & Destructor Documentation

rso::CStereoOdometryEstimator::TDetectParams::TDetectParams

(

)

Member Function Documentation

void rso::CStereoOdometryEstimator::TDetectParams::dumpToConsole ( )

inline

Member Data Documentation

TDMethod rso::CStereoOdometryEstimator::TDetectParams::detect_method

Method to detect features (also affects to the matching process)

int rso::CStereoOdometryEstimator::TDetectParams::fast_max_th

Limits for the FAST (within ORB) detector (dynamic) threshold.

int rso::CStereoOdometryEstimator::TDetectParams::fast_min_th

int rso::CStereoOdometryEstimator::TDetectParams::initial_FAST_threshold

The initial threshold of the FAST feature detector, which will be dynamically adapted to fulfill target_feats_per_pixel (Default=15)

int rso::CStereoOdometryEstimator::TDetectParams::KLT_win

Window for the KLT response evaluation (Default: 4)

size_t rso::CStereoOdometryEstimator::TDetectParams::min_distance

The allowed minimun distance between features.

double rso::CStereoOdometryEstimator::TDetectParams::minimum_KLT_response

Minimum KLT response to not to discard a feature as being in a textureless zone (Default: 10)

double rso::CStereoOdometryEstimator::TDetectParams::minimum_ORB_response

Minimum ORB response [Harris response] to not to discard a feature as being in a textureless zone (Default: 0.005)

NMSMethod rso::CStereoOdometryEstimator::TDetectParams::nmsMethod

Method to perform non maximal suppression.

bool rso::CStereoOdometryEstimator::TDetectParams::non_maximal_suppression

Enable/disable the non-maximal suppression after the detection (5x5 windows is used)

size_t rso::CStereoOdometryEstimator::TDetectParams::orb_nfeats

Number of features to be detected (only for ORB)

size_t rso::CStereoOdometryEstimator::TDetectParams::orb_nlevels

The number of pyramid levels.

double rso::CStereoOdometryEstimator::TDetectParams::target_feats_per_pixel

Desired number of features per square pixel (Default: 10/1000)

---

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

• libstereo-odometry.h