boofcv.abst.geo.calibration

Class CalibrateStereoPlanar

java.lang.Object

boofcv.abst.geo.calibration.CalibrateStereoPlanar

public class CalibrateStereoPlanar
extends java.lang.Object

Given a sequence of observations from a stereo camera compute the intrinsic calibration of each camera and the extrinsic calibration between the two cameras. A Planar calibration grid is used, which must be completely visible in all images.

Calibration is performed by first independently determining the intrinsic parameters of each camera as well as their extrinsic parameters relative to the calibration grid. Then the extrinsic parameters between the two cameras is found by creating two point clouds composed of the calibration points in each camera's view. Then the rigid body motion is found which transforms one point cloud into the other.

See comments in CalibrateMonoPlanar about when the y-axis should be inverted.

Constructor Summary

Constructors

Constructor and Description
CalibrateStereoPlanar(DetectorFiducialCalibration detector) Configures stereo calibration

Method Summary

Modifier and Type	Method and Description
boolean	addPair(boofcv.struct.image.GrayF32 left, boofcv.struct.image.GrayF32 right) Adds a pair of images that observed the same target.
void	configure(boolean assumeZeroSkew, int numRadialParam, boolean includeTangential) Specify calibration assumptions.
CalibrateMonoPlanar	getCalibLeft()
CalibrateMonoPlanar	getCalibRight()
void	printStatistics()
boofcv.struct.calib.StereoParameters	process() Compute stereo calibration parameters
void	reset() Puts the class into its initial state.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

CalibrateStereoPlanar

public CalibrateStereoPlanar(DetectorFiducialCalibration detector)

Configures stereo calibration

Parameters:

detector - Target detection algorithm.

Method Detail

reset

public void reset()

Puts the class into its initial state.

configure

public void configure(boolean assumeZeroSkew,
                      int numRadialParam,
                      boolean includeTangential)

Specify calibration assumptions.

Parameters:

assumeZeroSkew - If true zero skew is assumed.

numRadialParam - Number of radial parameters

includeTangential - If true it will estimate tangential distortion parameters.

addPair

public boolean addPair(boofcv.struct.image.GrayF32 left,
                       boofcv.struct.image.GrayF32 right)

Adds a pair of images that observed the same target.

Parameters:

left - Image of left target.

right - Image of right target.

process

public boofcv.struct.calib.StereoParameters process()

Compute stereo calibration parameters

Returns:

Stereo calibration parameters

getCalibLeft

public CalibrateMonoPlanar getCalibLeft()

getCalibRight

public CalibrateMonoPlanar getCalibRight()

printStatistics

public void printStatistics()