#include <plane_model_road_segmentation.h>

Public Member Functions
tf::Transform	find_and_publish_road_frame (pcl::PointCloud< T > cloud_in)
	get and publish the road frame T his f*unction does all the work for you. here are the main steps: STEP 1 create the model coeeficients STEP 2 project origin to plane, for example STEP 3 Find the difference between the cloud frame_id to ground STEP 4 Calculate the difference z axis STEP 5 Publish the road frame More...

tf::Transform	find_and_publish_road_frame (const sensor_msgs::PointCloud2ConstPtr &msg)
	overload function for ros msgs Converts sensor_msgs::PointCloud2 to pcl and calls the function find_and_publish_road_frame(pcl::PointCloud< T> pc) More...

bool	find_road_frame (pcl::PointCloud< T > cloud_in)
	get the road reference frame More...

bool	find_road_frame (const sensor_msgs::PointCloud2ConstPtr &msg)
	overload function for ros msgs Converts sensor_msgs::PointCloud2 to pcl and calls the function find_road_frame(pcl::PointCloud< T> pc) More...

pcl::PointCloud< T >	get_plane_coefficients (pcl::PointCloud< T > cloud_in, int iterations, float threshold, pcl::ModelCoefficients::Ptr coefficients)
	Calculate plane coefficients It uses pcl RANSAC model to estimate the coeeficients of a plane. More...

	plane_model_road ()

	~plane_model_road ()

Public Attributes
tf::TransformBroadcaster *	broadcast_ptr

pcl::ModelCoefficients::Ptr	coef

bool	debug

int	iterations

tf::TransformListener *	listener_atc_ground_ptr

T	point_origin

T	point_x

T	point_y

std::string	pub_frame_name

std::string	reference_ground_frame

float	threshold

Private Member Functions
bool	check_distance (double distance, double threshold)
	check if distance matches with threshold If returned true, you probably get the road plane More...

vector< double >	find_RP_angle (T origin_projected, T pt_x_projected, T pt_y_projected)
	find RP angles for road frame Gets the road frame RP angles. Notice that you don't need the Y. More...

tf::Transform	get_transform (T origin_projected, T pt_x_projected, T pt_y_projected)
	get the transformation point Gets the road transform to publish More...

T	project_point_to_plane (const T ptin, const pcl::ModelCoefficients::Ptr coeff)
	Project a point on a plane It uses pcl projectinliers to project a point on a plane. More...

T	transform_pcl_point (std::string frame1, std::string frame2, T pt_origin)
	Transform a pcl point between two ros frames It uses a tf::Stamped transform to transform a point between reference frames. More...

Private Attributes
T	origin_projected

pcl::PointCloud< T >	outliers

T	pt_x_projected

T	pt_y_projected

Detailed Description

template<class T>
class plane_model_road< T >

This is a class to detect the road reference frame

Date: March 2012

Author: ttalhada

Definition at line 72 of file plane_model_road_segmentation.h.

Constructor & Destructor Documentation

template<class T>

plane_model_road< T >::plane_model_road ( )

inline

Definition at line 148 of file plane_model_road_segmentation.h.

template<class T>

plane_model_road< T >::~plane_model_road ( )

inline

Definition at line 161 of file plane_model_road_segmentation.h.

Member Function Documentation

template<class T>

bool plane_model_road< T >::check_distance	(	double	distance,
		double	threshold
	)

inlineprivate

check if distance matches with threshold If returned true, you probably get the road plane

Parameters

[in]	distance
[in]	threshold

Returns: bool result

Definition at line 201 of file plane_model_road_segmentation.h.

template<class T >

tf::Transform plane_model_road< T >::find_and_publish_road_frame ( pcl::PointCloud< T > cloud_in )

get and publish the road frame T his f*unction does all the work for you. here are the main steps: STEP 1 create the model coeeficients STEP 2 project origin to plane, for example STEP 3 Find the difference between the cloud frame_id to ground STEP 4 Calculate the difference z axis STEP 5 Publish the road frame

Parameters

[in] origin_projected

Returns: tf::Transform

Definition at line 107 of file plane_model_road_segmentation_implementation.hpp.

template<class T >

tf::Transform plane_model_road< T >::find_and_publish_road_frame ( const sensor_msgs::PointCloud2ConstPtr & msg )

overload function for ros msgs Converts sensor_msgs::PointCloud2 to pcl and calls the function find_and_publish_road_frame(pcl::PointCloud< T> pc)

Parameters

[in] msg

Returns: transform

Definition at line 122 of file plane_model_road_segmentation_implementation.hpp.

template<class T >

bool plane_model_road< T >::find_road_frame ( pcl::PointCloud< T > cloud_in )

get the road reference frame

T h*is function does all the work for you. here are the main steps: STEP 1 create the model coeeficients STEP 2 project origin to plane, for example STEP 3 Find the difference between the cloud frame_id to ground STEP 4 Calculate the difference z axis

Parameters

[in] cloud_in

Returns: bool success

Definition at line 62 of file plane_model_road_segmentation_implementation.hpp.

template<class T >

bool plane_model_road< T >::find_road_frame ( const sensor_msgs::PointCloud2ConstPtr & msg )

overload function for ros msgs Converts sensor_msgs::PointCloud2 to pcl and calls the function find_road_frame(pcl::PointCloud< T> pc)

Parameters

[in] msg

Returns: success

Definition at line 94 of file plane_model_road_segmentation_implementation.hpp.

template<class T >

vector< double > plane_model_road< T >::find_RP_angle	(	T	origin_projected,
		T	pt_x_projected,
		T	pt_y_projected
	)

private

find RP angles for road frame Gets the road frame RP angles. Notice that you don't need the Y.

Parameters

[in]	origin_projected
[in]	pt_x_projected
[in]	pt_y_projected

Returns: std::vector< double> with R and P.

Definition at line 206 of file plane_model_road_segmentation_implementation.hpp.

template<class T >

pcl::PointCloud< T > plane_model_road< T >::get_plane_coefficients	(	pcl::PointCloud< T >	cloud_in,
		int	iterations,
		float	threshold,
		pcl::ModelCoefficients::Ptr	coefficients
	)