OCLSLAM< CR, CW > Class Template Reference

Interface class for the SLAM pipeline. More...

#include <ocl_processing.hpp>

Public Member Functions
	OCLSLAM (Kinect *kinect, octomap::OcTree &map)
	Constructor. More...
	~OCLSLAM ()
	Destructor. More...
void	init ()
	Initializes the SLAM pipeline. More...
void	registerPointCloud ()
	Registers a point cloud. More...
void	write (std::string filename=std::string("map.ot"))
	Stores an occupancy map on disk. More...
void	writeBinary (std::string filename=std::string("map.bt"))
	Stores an binary map on disk. More...
void	display ()
	Prints on the console results about the current registration and localization. More...
bool	getSLAMStatus ()
	Gets the status of the automated SLAM process. More...
void	setSLAMStatus (bool flag)
	Sets the status of the automated SLAM process. More...
void	toggleSLAMStatus ()
	Toggles the status of the automated SLAM process. More...
bool	getGFRGBStatus ()
	Gets the status of the RGB Guided Filter. More...
void	setGFRGBStatus (bool flag)
	Sets the status of the RGB Guided Filter. More...
void	toggleGFRGBStatus ()
	Toggles the status of the RGB Guided Filter. More...
bool	getGFDStatus ()
	Gets the status of the Depth Guided Filter. More...
void	setGFDStatus (bool flag)
	Sets the status of the Depth Guided Filter. More...
void	toggleGFDStatus ()
	Toggles the status of the Depth Guided Filter. More...
int	getRGBNormalization ()
	Gets the status of the RGB normalization. More...
void	setRGBNormalization (int flag)
	Sets the status of the RGB normalization. More...
void	toggleRGBNormalization ()
	Toggles the status of the RGB normalization. More...
int	getGFRGBRadius ()
	Gets the window radius \(r\) for the guided filter performed on the RGB frame. More...
void	setGFRGBRadius (int radius)
	Sets the window radius \(r\) for the guided filter performed on the RGB frame. More...
float	getGFRGBEps ()
	Gets the variability threshold \(\epsilon\) for the guided filter performed on the RGB frame. More...
void	setGFRGBEps (float eps)
	Sets the variability threshold \(\epsilon\) for the guided filter performed on the RGB frame. More...
int	getGFDRadius ()
	Gets the window radius \(r\) for the guided filter performed on the Depth frame. More...
void	setGFDRadius (int radius)
	Sets the window radius \(r\) for the guided filter performed on the Depth frame. More...
float	getGFDEps ()
	Gets the variability threshold \(\epsilon\) for the guided filter performed on the Depth frame. More...
void	setGFDEps (float eps)
	Sets the variability threshold \(\epsilon\) for the guided filter performed on the Depth frame. More...
float	getGFDScaling ()
	Gets the scaling applied to the depth frame for processing with the guided filter. More...
void	setGFDScaling (float scaling)
	Sets the scaling applied to the depth frame for processing with the guided filter. More...
float	getSensorFocalLength ()
	Gets the sensor's focal length. More...
void	setSensorFocalLength (float f)
	Sets the sensor's focal length. More...
float	getRBCAlpha ()
	Gets the parameter \( \alpha \) used in the distance function for the RBC data structure. More...
void	setRBCAlpha (float _a)
	Sets the parameter \( \alpha \) used in the distance function for the RBC data structure. More...
float	getICPSScaling ()
	Gets the scaling applying to the deviations when computing matrix S in the ICP algorithm, for managing floating point arithmetic issues. More...
void	setICPSScaling (float _c)
	Sets the scaling applying to the deviations when computing matrix S in the ICP algorithm, for managing floating point arithmetic issues. More...
unsigned int	getICPMaxIterations ()
	Gets the maximum number of iterations considered for an ICP registration. More...
void	setICPMaxIterations (unsigned int maxIter)
	Sets the maximum number of iterations considered for an ICP registration. More...
double	getICPAngleThreshold ()
	Gets the angle threshold (in degrees) for the convergence check of the ICP. More...
void	setICPAngleThreshold (double at)
	Sets the angle threshold (in degrees) for the convergence check of the ICP. More...
double	getICPTranslationThreshold ()
	Gets the translation threshold (in mm) for the convergence check of the ICP. More...
void	setICPTranslationThreshold (double tt)
	Sets the translation threshold (in mm) for the convergence check of the ICP. More...

Public Attributes
std::function< void()>	slam
	Performs the SLAM process. More...
volatile int	timeStep
Eigen::Matrix3f	R_g
Eigen::Quaternionf	q_g
Eigen::Vector3f	t_g
cl_float	s_g

Detailed Description

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>
class OCLSLAM< CR, CW >

Interface class for the SLAM pipeline.

Retrieves data, registers point clouds, and builds a map.

Note

In order to handle memory consumption, there is a limit on the number of point clouds displayed on the screen. When the limit is reached though, the SLAM process keeps running on the background and the map gets updated as normal.

Template Parameters

CR	configures the class with different methods of rotation computation.
CW	configures the class for performing either regular or weighted computation.

Constructor & Destructor Documentation

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

OCLSLAM< CR, CW >::OCLSLAM	(	Kinect *	kinect,
		octomap::OcTree &	map
	)

Constructor.

Initializes the OpenCL environment, the OpenGL buffers, and the classes for the SLAM pipeline.

Parameters

[in]	kinect	initialized Kinect device.
[in]	map	OctoMap structure for building the map.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

OCLSLAM< CR, CW >::~OCLSLAM

(

)

Destructor.

Member Function Documentation

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::display

(

)

Prints on the console results about the current registration and localization.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

float OCLSLAM< CR, CW >::getGFDEps ( )

inline

Gets the variability threshold \(\epsilon\) for the guided filter performed on the Depth frame.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

int OCLSLAM< CR, CW >::getGFDRadius ( )

inline

Gets the window radius \(r\) for the guided filter performed on the Depth frame.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

float OCLSLAM< CR, CW >::getGFDScaling ( )

inline

Gets the scaling applied to the depth frame for processing with the guided filter.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

bool OCLSLAM< CR, CW >::getGFDStatus ( )

inline

Gets the status of the Depth Guided Filter.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

float OCLSLAM< CR, CW >::getGFRGBEps ( )

inline

Gets the variability threshold \(\epsilon\) for the guided filter performed on the RGB frame.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

int OCLSLAM< CR, CW >::getGFRGBRadius ( )

inline

Gets the window radius \(r\) for the guided filter performed on the RGB frame.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

bool OCLSLAM< CR, CW >::getGFRGBStatus ( )

inline

Gets the status of the RGB Guided Filter.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

double OCLSLAM< CR, CW >::getICPAngleThreshold ( )

inline

Gets the angle threshold (in degrees) for the convergence check of the ICP.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

unsigned int OCLSLAM< CR, CW >::getICPMaxIterations ( )

inline

Gets the maximum number of iterations considered for an ICP registration.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

float OCLSLAM< CR, CW >::getICPSScaling ( )

inline

Gets the scaling applying to the deviations when computing matrix S in the ICP algorithm, for managing floating point arithmetic issues.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

double OCLSLAM< CR, CW >::getICPTranslationThreshold ( )

inline

Gets the translation threshold (in mm) for the convergence check of the ICP.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

float OCLSLAM< CR, CW >::getRBCAlpha ( )

inline

Gets the parameter \( \alpha \) used in the distance function for the RBC data structure.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

int OCLSLAM< CR, CW >::getRGBNormalization ( )

inline

Gets the status of the RGB normalization.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

float OCLSLAM< CR, CW >::getSensorFocalLength ( )

inline

Gets the sensor's focal length.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

bool OCLSLAM< CR, CW >::getSLAMStatus ( )

inline

Gets the status of the automated SLAM process.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::init

(

)

Initializes the SLAM pipeline.

It sets up the first point cloud. After that, registerPointCloud can be called to perform the registration process.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::registerPointCloud

(

)

Registers a point cloud.

Runs ICP on the current point cloud, delivers the result to OpenGL for visualization, and to OctoMap, in order to update the map.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setGFDEps ( float  eps )

inline

Sets the variability threshold \(\epsilon\) for the guided filter performed on the Depth frame.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setGFDRadius ( int  radius )

inline

Sets the window radius \(r\) for the guided filter performed on the Depth frame.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setGFDScaling ( float  scaling )

inline

Sets the scaling applied to the depth frame for processing with the guided filter.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setGFDStatus ( bool  flag )

inline

Sets the status of the Depth Guided Filter.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setGFRGBEps ( float  eps )

inline

Sets the variability threshold \(\epsilon\) for the guided filter performed on the RGB frame.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setGFRGBRadius ( int  radius )

inline

Sets the window radius \(r\) for the guided filter performed on the RGB frame.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setGFRGBStatus ( bool  flag )

inline

Sets the status of the RGB Guided Filter.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setICPAngleThreshold ( double  at )

inline

Sets the angle threshold (in degrees) for the convergence check of the ICP.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setICPMaxIterations ( unsigned int  maxIter )

inline

Sets the maximum number of iterations considered for an ICP registration.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setICPSScaling ( float  _c )

inline

Sets the scaling applying to the deviations when computing matrix S in the ICP algorithm, for managing floating point arithmetic issues.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setICPTranslationThreshold ( double  tt )

inline

Sets the translation threshold (in mm) for the convergence check of the ICP.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setRBCAlpha ( float  _a )

inline

Sets the parameter \( \alpha \) used in the distance function for the RBC data structure.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setRGBNormalization ( int  flag )

inline

Sets the status of the RGB normalization.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setSensorFocalLength ( float  f )

inline

Sets the sensor's focal length.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::setSLAMStatus

(

bool

flag

)

Sets the status of the automated SLAM process.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::toggleGFDStatus ( )

inline

Toggles the status of the Depth Guided Filter.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::toggleGFRGBStatus ( )

inline

Toggles the status of the RGB Guided Filter.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::toggleRGBNormalization ( )

inline

Toggles the status of the RGB normalization.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::toggleSLAMStatus

(

)

Toggles the status of the automated SLAM process.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::write

(

std::string

filename = std::string ("map.ot")

)

Stores an occupancy map on disk.

Parameters

[in]

filename

name for the map file [.ot].

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

void OCLSLAM< CR, CW >::writeBinary

(

std::string

filename = std::string ("map.bt")

)

Stores an binary map on disk.

Parameters

[in]

filename

name for the map file [.bt].

Member Data Documentation

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

Eigen::Quaternionf OCLSLAM< CR, CW >::q_g

Represents the orietation with respect to the global coordinate frame, given in quaternion representation, \( \dot{q}_g = \left[ \begin{matrix} q_x & q_y & q_z & q_w \end{matrix} \right]^T \).

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

Eigen::Matrix3f OCLSLAM< CR, CW >::R_g

Represents the orietation with respect to the global coordinate frame, given in rotation matrix representation, \( R_g \).

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

cl_float OCLSLAM< CR, CW >::s_g

Represents the scale of the current point cloud with respect to the first one, given as a scalar, \( s_g \). It's used when transforming the current point cloud before mapping.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

std::function<void ()> OCLSLAM< CR, CW >::slam

Performs the SLAM process.

Initially, it points to init for registering the first point cloud, and after that it gets assigned with a function that runs repeatedly registerPointCloud.

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

Eigen::Vector3f OCLSLAM< CR, CW >::t_g

Represents the translation (in mm) with respect to the global coordinate frame, given as a vector in 3-D, \( t_g \).

template<ICP::ICPStepConfigT CR, ICP::ICPStepConfigW CW>

volatile int OCLSLAM< CR, CW >::timeStep

Counts the discrete time steps, as in the number of point clouds registered.

---

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

• ocl_processing.hpp
• ocl_processing.cpp