rso::CStereoOdometryEstimator::TLeastSquaresParams Struct Reference

#include <libstereo-odometry.h>

Public Member Functions
	TLeastSquaresParams ()
void	dumpToConsole ()

Public Attributes
bool	use_robust_kernel
	(def:true) – Set/Unset using robust kernel for optimization More...
double	kernel_param
	(def:3) – Robust kernel parameter (pseudo-Huber) More...
size_t	max_iters
	(def:100) – Final maximum number of iterations for the refinement stage More...
size_t	initial_max_iters
	(def:10) – Maximum number of iterations for the initial stage More...
double	min_mod_out_vector
	(def:0.001) – Minimum modulus of the step output vector to continue iterating (ending condition) More...
double	std_noise_pixels
	(def:1) – The standard deviation assumed for feature coordinates (this parameter is only needed to scale the uncertainties of reconstructed LMs with unknown locations). (unused by now) More...
size_t	max_incr_cost
	(def:3) – Maximum allowed number of times the cost can grow More...
double	residual_threshold
	(def:10) – Residual threshold for detecting outliers More...
size_t	bad_tracking_th
	(def:5) – Minimum number of tracked features to yield a tracking error More...
bool	use_previous_pose_as_initial
	(def:true) – Use the previous computed pose as the initial one for the next frame More...
bool	use_custom_initial_pose
	(def:false) – Use the (input) custom initial pose but do not save it for later use (useful when getChangeInPose is called). This setting has priority over 'use_previous_pose_as_initial' More...

Detailed Description

Parameters for the LS optimization stage

Constructor & Destructor Documentation

rso::CStereoOdometryEstimator::TLeastSquaresParams::TLeastSquaresParams

(

)

Member Function Documentation

void rso::CStereoOdometryEstimator::TLeastSquaresParams::dumpToConsole ( )

inline

Member Data Documentation

size_t rso::CStereoOdometryEstimator::TLeastSquaresParams::bad_tracking_th

(def:5) – Minimum number of tracked features to yield a tracking error

size_t rso::CStereoOdometryEstimator::TLeastSquaresParams::initial_max_iters

(def:10) – Maximum number of iterations for the initial stage

double rso::CStereoOdometryEstimator::TLeastSquaresParams::kernel_param

(def:3) – Robust kernel parameter (pseudo-Huber)

size_t rso::CStereoOdometryEstimator::TLeastSquaresParams::max_incr_cost

(def:3) – Maximum allowed number of times the cost can grow

size_t rso::CStereoOdometryEstimator::TLeastSquaresParams::max_iters

(def:100) – Final maximum number of iterations for the refinement stage

double rso::CStereoOdometryEstimator::TLeastSquaresParams::min_mod_out_vector

(def:0.001) – Minimum modulus of the step output vector to continue iterating (ending condition)

double rso::CStereoOdometryEstimator::TLeastSquaresParams::residual_threshold

(def:10) – Residual threshold for detecting outliers

double rso::CStereoOdometryEstimator::TLeastSquaresParams::std_noise_pixels

(def:1) – The standard deviation assumed for feature coordinates (this parameter is only needed to scale the uncertainties of reconstructed LMs with unknown locations). (unused by now)

bool rso::CStereoOdometryEstimator::TLeastSquaresParams::use_custom_initial_pose

(def:false) – Use the (input) custom initial pose but do not save it for later use (useful when getChangeInPose is called). This setting has priority over 'use_previous_pose_as_initial'

bool rso::CStereoOdometryEstimator::TLeastSquaresParams::use_previous_pose_as_initial

(def:true) – Use the previous computed pose as the initial one for the next frame

bool rso::CStereoOdometryEstimator::TLeastSquaresParams::use_robust_kernel

(def:true) – Set/Unset using robust kernel for optimization

---

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

• libstereo-odometry.h