Camera class holds a set of camera parameters together with the window pos/size. More...

#include <pcl/visualization/common/common.h>

Public Member Functions
void	computeViewMatrix (Eigen::Matrix4d &view_mat) const
	Computes View matrix for Camera (Based on gluLookAt)
void	computeProjectionMatrix (Eigen::Matrix4d &proj) const
	Computes Projection Matrix for Camera.
template<typename PointT >
void	cvtWindowCoordinates (const PointT &pt, Eigen::Vector4d &window_cord)
	converts point to window coordiantes
template<typename PointT >
void	cvtWindowCoordinates (const PointT &pt, Eigen::Vector4d &window_cord, const Eigen::Matrix4d &composite_mat)
	converts point to window coordiantes
Public Attributes
double	focal [3]
double	pos [3]
double	view [3]
double	clip [2]
double	fovy
double	window_size [2]
double	window_pos [2]

Detailed Description

Camera class holds a set of camera parameters together with the window pos/size.

Definition at line 79 of file common.h.

Member Function Documentation

void pcl::visualization::Camera::computeProjectionMatrix ( Eigen::Matrix4d & proj ) const

Computes Projection Matrix for Camera.

Parameters:

[out] proj the resultant matrix

void pcl::visualization::Camera::computeViewMatrix ( Eigen::Matrix4d & view_mat ) const

Computes View matrix for Camera (Based on gluLookAt)

Parameters:

[out] view_mat the resultant matrix

template<typename PointT >

void pcl::visualization::Camera::cvtWindowCoordinates	(	const PointT &	pt,
		Eigen::Vector4d &	window_cord
	)

converts point to window coordiantes

Parameters:

[in]	pt	xyz point to be converted
[out]	window_cord	vector containing the pts' window X,Y, Z and 1

This function computes the projection and view matrix every time. It is very inefficient to use this for every point in the point cloud!

template<typename PointT >

void pcl::visualization::Camera::cvtWindowCoordinates	(	const PointT &	pt,
		Eigen::Vector4d &	window_cord,
		const Eigen::Matrix4d &	composite_mat
	)

converts point to window coordiantes

Parameters:

[in]	pt	xyz point to be converted
[out]	window_cord	vector containing the pts' window X,Y, Z and 1
[in]	composite_mat	composite transformation matrix (proj*view)

Use this function to compute window coordinates with a precomputed transformation function. The typical composite matrix will be the projection matrix * the view matrix. However, additional matrices like a camera disortion matrix can also be added.