GTSAM  4.0.2
C++ library for smoothing and mapping (SAM)
Public Types | Protected Attributes | List of all members
gtsam::Cal3DS2_Base Class Reference

Calibration of a camera with radial distortion. More...

#include <Cal3DS2_Base.h>

Inheritance diagram for gtsam::Cal3DS2_Base:
Inheritance graph
[legend]
Collaboration diagram for gtsam::Cal3DS2_Base:
Collaboration graph
[legend]

Public Types

enum  { dimension = 9 }
 
using shared_ptr = std::shared_ptr< Cal3DS2_Base >
 
enum  { dimension = 5 }
 

Protected Attributes

double k1_ = 0.0f
 
double k2_ = 0.0f
 radial 2nd-order and 4th-order
 
double p1_ = 0.0f
 
double p2_ = 0.0f
 tangential distortion
 
double tol_ = 1e-5
 tolerance value when calibrating
 
double fx_ = 1.0f
 
double fy_ = 1.0f
 focal length
 
double s_ = 0.0f
 skew
 
double u0_ = 0.0f
 
double v0_ = 0.0f
 principal point
 

Testable

bool equals (const Cal3 &K, double tol=10e-9) const
 Check if equal up to specified tolerance.
 

Standard Interface

double fx () const
 focal length x
 
double fy () const
 focal length y
 
double aspectRatio () const
 aspect ratio
 
double skew () const
 skew
 
double px () const
 image center in x
 
double py () const
 image center in y
 
Point2 principalPoint () const
 return the principal point
 
virtual Matrix3 K () const
 return calibration matrix K
 
Matrix3 inverse () const
 Return inverted calibration matrix inv(K)
 

Standard Constructors

 Cal3DS2_Base ()=default
 Default Constructor with only unit focal length.
 
 Cal3DS2_Base (double fx, double fy, double s, double u0, double v0, double k1, double k2, double p1=0.0, double p2=0.0, double tol=1e-5)
 
 ~Cal3DS2_Base () override
 

Advanced Constructors

 Cal3DS2_Base (const Vector9 &v)
 

Testable

void print (const std::string &s="") const override
 print with optional string
 
bool equals (const Cal3DS2_Base &K, double tol=1e-8) const
 assert equality up to a tolerance
 
GTSAM_EXPORT friend std::ostream & operator<< (std::ostream &os, const Cal3DS2_Base &cal)
 Output stream operator.
 

Standard Interface

double k1 () const
 First distortion coefficient.
 
double k2 () const
 Second distortion coefficient.
 
double p1 () const
 First tangential distortion coefficient.
 
double p2 () const
 Second tangential distortion coefficient.
 
Vector4 k () const
 return distortion parameter vector
 
Vector9 vector () const
 Return all parameters as a vector.
 
Point2 uncalibrate (const Point2 &p, OptionalJacobian< 2, 9 > Dcal={}, OptionalJacobian< 2, 2 > Dp={}) const
 
Point2 calibrate (const Point2 &p, OptionalJacobian< 2, 9 > Dcal={}, OptionalJacobian< 2, 2 > Dp={}) const
 Convert (distorted) image coordinates uv to intrinsic coordinates xy.
 
Matrix2 D2d_intrinsic (const Point2 &p) const
 Derivative of uncalibrate wrpt intrinsic coordinates.
 
Matrix29 D2d_calibration (const Point2 &p) const
 Derivative of uncalibrate wrpt the calibration parameters.
 
size_t dim () const override
 return DOF, dimensionality of tangent space
 
static size_t Dim ()
 return DOF, dimensionality of tangent space
 

Clone

virtual std::shared_ptr< Cal3DS2_Baseclone () const
 

Detailed Description

Calibration of a camera with radial distortion.

Uses same distortionmodel as OpenCV, with http://docs.opencv.org/modules/calib3d/doc/camera_calibration_and_3d_reconstruction.html but using only k1,k2,p1, and p2 coefficients. K = [ fx s u0 ; 0 fy v0 ; 0 0 1 ] r² = P.x² + P.y² P̂ = (1 + k1*r² + k2*r⁴) P + [ (2*p1 P.x P.y) + p2 (r² + 2 Pn.x²) p1 (r² + 2 Pn.y²) + (2*p2 Pn.x Pn.y) ] pi = K*P̂

Member Function Documentation

◆ clone()

virtual std::shared_ptr<Cal3DS2_Base> gtsam::Cal3DS2_Base::clone ( ) const
inlinevirtual
Returns
a deep copy of this object

Reimplemented in gtsam::Cal3DS2.

◆ uncalibrate()

Point2 gtsam::Cal3DS2_Base::uncalibrate ( const Point2 p,
OptionalJacobian< 2, 9 >  Dcal = {},
OptionalJacobian< 2, 2 >  Dp = {} 
) const

convert intrinsic coordinates xy to (distorted) image coordinates uv

Parameters
ppoint in intrinsic coordinates
Dcaloptional 2*9 Jacobian wrpt Cal3DS2 parameters
Dpoptional 2*2 Jacobian wrpt intrinsic coordinates
Returns
point in (distorted) image coordinates

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