Test Calibration

[teganscott]

Is there a method I can use to test the calibration? When I put the calibration numbers into my own software (which works with opencv and calibrations from other systems) it doesn't coorrelate the camera and projector correctly.

Thank you,
Tegan

[saudet]

Could you copy/paste typical results that you get here?

[teganscott]

Below is what procamlib gives me, and the formatting I have for my calibration files. These two will be slightly different because the projector and camera are in different positions. The biggest difference is in the extrinsic rotation of my old calibrations was a 3x1 matrix, and the new one is 3x3. In my code I have the following that converts the rotation matrix into Rodrigues:

cv::Vec3d axis_angle;
double data[9] = {9.9981890125721906e-01, -8.7052645857853533e-03,
1.6922856062103579e-02, 1.6082024959920501e-02,
8.6194239428941977e-01, -5.0675110004795354e-01,
-1.0175124667329560e-02, 5.0693148185441583e-01,
8.6192629588781566e-01};
cv::Mat origRot = cv::Mat(3,3, CV_64F, data);
cv::Rodrigues(origRot, axis_angle);
cv::Mat rot = cv::Mat(3, 1, CV_64F, axis_angle);

This is what procamlib gives me:

%YAML:1.0

Cameras:

• Camera 0
  Projectors:
• Projector 0
  Camera 0:
  imageWidth: 640
  imageHeight: 480
  responseGamma: 0.
  cameraMatrix: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ 7.7581750624364702e+02, 0., 3.2712277369620023e+02, 0.,
  7.7610945139991838e+02, 2.3484626839931556e+02, 0., 0., 1. ]
  distortionCoeffs: !!opencv-matrix
  rows: 1
  cols: 4
  dt: d
  data: [ -1.1150538574733834e-01, 1.7528485231520349e-01,
  1.3777229721069281e-04, 3.5512804807152055e-03 ]
  extrParams: !!opencv-matrix
  rows: 10
  cols: 6
  dt: d
  data: [ -2.4156499206165061e-01, 1.8012775548542140e-01,
  3.0971126521233794e+00, 9.7672612756320257e+02,
  8.4309337872134495e+02, 3.0579462522394706e+03,
  -7.8549242710261080e-01, -2.8501749151166361e-01,
  -2.9269460537869385e+00, 8.6784058799846707e+02,
  8.7237706917774949e+02, 2.5111457941924082e+03,
  1.7173840480802446e-01, 7.6806524901152817e-01,
  -3.0011869667530067e+00, 8.1736575306125781e+02,
  5.9774869098573743e+02, 3.1628262239991532e+03,
  -4.9985867968159597e-02, -9.3919224795251288e-01,
  -2.8871041470444330e+00, 1.1068227286276594e+03,
  1.0505865212093518e+03, 3.0583236718773715e+03,
  1.2579745239764833e-01, 9.2067055083533952e-02,
  -3.1256843279217237e+00, 9.0416180531444070e+02,
  7.0449061666552848e+02, 2.5043720925455700e+03,
  2.7428722866650290e-01, 8.7245690431954001e-02,
  3.1078726288431393e+00, 6.8503074658642504e+02,
  1.0020846281253456e+03, 4.1669747122728004e+03,
  -1.1503040335019099e+00, 1.7024453055326852e-01,
  2.8854476605390929e+00, 4.5169684320598648e+01,
  1.1418006977246384e+03, 5.0062320468696498e+03,
  -1.2036048151279497e+00, 9.0617781989513636e-04,
  -2.7793536264733043e+00, 4.9855943180165428e+02,
  9.8652834329722725e+02, 3.5770730312324645e+03,
  3.9387072333177947e-01, -1.3150287815934847e-01,
  -3.0678627249891148e+00, 3.5969986263255259e+02,
  1.1399110179951804e+03, 5.0285896568295348e+03,
  3.1649409187408778e-01, -1.4772840685344468e-01,
  -3.0460147144952132e+00, 1.3429765374386300e+03,
  1.1414031663775518e+03, 3.4867023648561662e+03 ]
  reprojErrs: !!opencv-matrix
  rows: 1
  cols: 10
  dt: d
  data: [ 1.5371241343872963e-01, 2.8960308273486646e-01,
  2.8750924207329859e-01, 2.6344386342339299e-01,
  1.9743376450287983e-01, 1.4753773852132168e-01,
  2.9277691879812734e-01, 2.3886584943637879e-01,
  1.3513256849293118e-01, 2.4715896890807423e-01 ]
  avgReprojErr: 2.2898737268678926e-01
  maxReprojErr: 5.5011680951627251e-01
  R: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ 1., 0., 0., 0., 1., 0., 0., 0., 1. ]
  T: !!opencv-matrix
  rows: 3
  cols: 1
  dt: d
  data: [ 0., 0., 0. ]
  E: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ 0., 0., 0., 0., 0., 0., 0., 0., 0. ]
  F: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ 0., 0., 0., 0., 0., 0., 0., 0., 0. ]
  avgEpipolarErr: 0.
  maxEpipolarErr: 0.
  colorOrder: BGR
  colorMixingMatrix: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ 1., 0., 0., 0., 1., 0., 0., 0., 1. ]
  additiveLight: !!opencv-matrix
  rows: 3
  cols: 1
  dt: d
  data: [ 0., 0., 0. ]
  avgColorErr: 0.
  colorR2: 1.
  Projector 0:
  imageWidth: 1280
  imageHeight: 720
  responseGamma: 2.2000000000000002e+00
  cameraMatrix: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ 1.7499631210240912e+03, 0., 6.1660168685378312e+02, 0.,
  1.6896417183771678e+03, 5.9724917683269268e+02, 0., 0., 1. ]
  distortionCoeffs: !!opencv-matrix
  rows: 1
  cols: 4
  dt: d
  data: [ -6.6930274333599230e-02, -6.7048107227998741e-03,
  3.5010575123940813e-03, -4.7314086914342478e-03 ]
  extrParams: !!opencv-matrix
  rows: 10
  cols: 6
  dt: d
  data: [ -1.9354736345635731e-01, -6.1930165346661692e-01,
  3.0187165004280700e+00, 8.5046734215751610e+02,
  7.4581100284211914e+01, 3.8326438369483453e+03,
  -7.0076107953804545e-01, 4.5589671940981369e-01,
  -2.7754234848766712e+00, 7.3094205313373061e+02,
  3.6713582005614649e+02, 3.3942323715873163e+03,
  1.4541709844879613e-01, 1.5413076626580589e+00,
  -2.7117090000409840e+00, 6.9619102699464202e+02,
  -1.9936683545410492e+02, 3.8187930255447154e+03,
  -3.3716930446300100e-02, -1.4679683230627441e-01,
  -3.0091628510661144e+00, 9.7759350382268417e+02,
  2.5685067132913179e+02, 3.9146135814326717e+03,
  9.0205361217915572e-02, 8.6954962865644581e-01,
  -3.0120011161430216e+00, 7.6866096349570864e+02,
  2.3546879754297171e+02, 3.3043057744979283e+03,
  -2.6449754127169411e-01, 7.7466813197389139e-01,
  -2.9912285437944126e+00, 5.8054143325837515e+02,
  -3.5399942533497466e+02, 4.8746897121799084e+03,
  -1.0018636253415676e+00, -5.4434814044426161e-01,
  2.7031010656401122e+00, -4.3939248233586440e+01,
  -6.5648750361818213e+02, 5.6725145820799644e+03,
  -1.0579535623781158e+00, 6.7401045910367141e-01,
  -2.5168695575013897e+00, 3.8391487256352377e+02,
  -7.4323400666258678e+01, 4.3831002089077683e+03,
  3.2410786388553042e-01, 7.7752121371865823e-01,
  -3.0249338868293578e+00, 2.6858102654901393e+02,
  -6.6338509230912939e+02, 5.6337688820228377e+03,
  3.2017570407206858e-01, 6.6504050110457080e-01,
  -3.0195212490458379e+00, 1.2211691022937016e+03,
  1.2375505861043740e+02, 4.3269710530395978e+03 ]
  reprojErrs: !!opencv-matrix
  rows: 1
  cols: 10
  dt: d
  data: [ 1.3164197849823371e+00, 1.3514771113547357e+00,
  6.5346257455760870e-01, 1.1105567453156828e+00,
  1.5099777827348959e+00, 9.4673348127230028e-01,
  8.6161959791194154e-01, 9.8421655271010278e-01,
  1.0521663658108733e+00, 1.0783660369344115e+00 ]
  avgReprojErr: 1.1179781153666539e+00
  maxReprojErr: 2.6356199800838387e+00
  R: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ 9.9981890125721906e-01, -8.7052645857853533e-03,
  1.6922856062103579e-02, 1.6082024959920501e-02,
  8.6194239428941977e-01, -5.0675110004795354e-01,
  -1.0175124667329560e-02, 5.0693148185441583e-01,
  8.6192629588781566e-01 ]
  T: !!opencv-matrix
  rows: 3
  cols: 1
  dt: d
  data: [ -1.7107953358194348e+02, 8.7837854385005460e+02,
  7.8703985608089772e+02 ]
  E: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ -2.1594805798726945e+01, -2.3310528108853714e+02,
  1.1559308775385889e+03, 7.8515656857021804e+02,
  7.9874211286914274e+01, 1.6077691088209545e+02,
  -8.8097077592927394e+02, -1.3981418515884596e+02,
  7.1830068172726044e+01 ]
  F: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ -1.3368997344057221e-07, -1.4425744961417022e-06,
  5.9344078247353105e-03, 5.0343113925375577e-06,
  5.1194934569914426e-07, -9.6729370437643948e-04,
  -1.2468513687286359e-02, -9.3040828326544395e-04, 1. ]
  avgEpipolarErr: 2.6198481877793398e+00
  maxEpipolarErr: 7.0856579958364510e+00
  colorOrder: BGR
  colorMixingMatrix: !!opencv-matrix
  rows: 3
  cols: 3
  dt: d
  data: [ 9.6289023041437521e-02, 1.5850582325576001e-01,
  8.4659356542572395e-02, 5.1802646787281817e-03,
  2.3662243713180331e-01, 1.4141216619711117e-01,
  2.5245874910261937e-03, 4.7465533462665924e-02,
  3.3407212796649421e-01 ]
  additiveLight: !!opencv-matrix
  rows: 3
  cols: 1
  dt: d
  data: [ 4.9442887747905900e-01, 4.4000607242526013e-01,
  4.3617957606834917e-01 ]
  avgColorErr: 7.2630983371695351e-02
  colorR2: 6.7542116700531341e-01

This is what the calibration software I am working with needs:
CAMERA:
%YAML:1.0
cameraMatrix: !!opencv-matrix
rows: 3
cols: 3
dt: d
data: [ 5.2390037217688837e+02, 0., 2.9894316833599152e+02, 0.,
5.2273440154265677e+02, 2.3062940702516227e+02, 0., 0., 1. ]
imageSize_width: 640
imageSize_height: 480
sensorSize_width: 0
sensorSize_height: 0
distCoeffs: !!opencv-matrix
rows: 5
cols: 1
dt: d
data: [ -9.4425188361295043e-02, 1.3559756613205826e-01,
2.5196205369225116e-04, -2.2982133039631367e-03,
2.9994602486057755e-02 ]
reprojectionError: 1.5748588740825653e-01

PROJECTOR:
%YAML:1.0
cameraMatrix: !!opencv-matrix
rows: 3
cols: 3
dt: d
data: [ 2.0071611042360921e+03, 0., 6.2447724667816487e+02, 0.,
2.0000916462185571e+03, 1.5582659966238060e+02, 0., 0., 1. ]
imageSize_width: 1280
imageSize_height: 720
sensorSize_width: 0
sensorSize_height: 0
distCoeffs: !!opencv-matrix
rows: 5
cols: 1
dt: d
data: [ -9.8245651394649747e-02, 5.6098143869503347e-01,
1.1589744437146346e-02, 4.8286095453437001e-03,
5.6681736350551748e-01 ]
reprojectionError: 5.2733808755874634e-01

EXTRINSICS:
%YAML:1.0
Rotation_Vector: !!opencv-matrix
rows: 3
cols: 1
dt: d
data: [ 6.8210825021344387e-02, 2.8757581667304671e-02,
-3.0531880320480636e-02 ]
Translation_Vector: !!opencv-matrix
rows: 3
cols: 1
dt: d
data: [ 2.4228539839093036e-01, 3.1220677070266731e+00,
2.1248321280799489e+00 ]

[saudet]

Good! The errors don't look too big, so the calibration is OK. You'll just need to plug in the numbers correctly, yes.