cv2.CreateMat in python

edit

in cv2, everything's a numpy array. ( it does not need any Create* functions anymore. if you ask me - for the better! )

Sorry but i didn't understand you (i'm new in opencv and python) how can i create matrix?!

>>> import numpy as np

>>> a = numpy.array((3,4,3), np.uint8) # (that is: height, width,numchannels)

(i'm going to try a more detailled answer tomorrow ... )

okay thank you so much (please if you can help me in cv.stereoCalibration, how to calculate R,T,F,E?i'll be thankful) waiting for your explanation tomorrow.

use cv2.stereoCalibrate - it will just return all of them.

in general, stick to the cv2 api. the old cv one won't be included in the next release

ok,i did it and it works now thank you one more question please what will the next step be in order to be able to calculate the distance between the tow webcams and the object in front of them?

have a look at StereoBM or StereoSGBM to create a disparity map, which is (inverse)proportional to the distance. maybe this sample is helpful

okay working on it thank you

hi again, i tried to run the cv2.stereoRectify but i got this error:

Traceback (most recent call last): File "C:\opencv246\samples\python2\stereocalib.py", line 132, in <module> cv.stereoRectify(camera_matrixL, cameraR_matrix, dist_coefsL, dist_coefsR,(384,288), R, T, RL, Rr, PL, PR) error: ......\src\opencv\modules\imgproc\src\undistort.cpp:292: error: (-215) CV_IS_MAT(_distCoeffs) && (_distCoeffs->rows == 1 || _distCoeffs->cols == 1) && (_distCoeffs->rows_distCoeffs->cols == 4 || _distCoeffs->rows_distCoeffs->cols == 5 || _distCoeffs->rows*_distCoeffs->cols == 8) why?!

please use: cv2.stereoRectify(cameraMatrix1, distCoeffs1, cameraMatrix2, distCoeffs2, image, ...

imho, you used the cv one, and mixed up the arguments

python has a nice help function, btw:

>>> import cv2

>>> help(cv2.stereoRectify)

Help on built-in function stereoRectify in module cv2:

stereoRectify(...) stereoRectify(cameraMatrix1, distCoeffs1, cameraMatrix2, distCoeffs2, imageS ize, R, T[, R1[, R2[, P1[, P2[, Q[, flags[, alpha[, newImageSize]]]]]]]]) -> R1, R2, P1, P2, Q, validPixROI1, validPixROI2

yes it is!

i am using cv2 but just (import cv2 as cv)

and how about the R1,R2,P1,P2,Q?

"and how about the R1,R2,P1,P2,Q?" - pardon ? they'll get returned from the function. are you asking, wether you have to supply them as arguments ? - no. (well you could ofc, but that's optional)

ok,but still getting this error: ValueError: too many values to unpack in this line RL, Rr, PL, PR, Q = cv.stereoRectify(camera_matrixL,dist_coefsL, cameraR_matrix, dist_coefsR,(384,288), R, T) (and i'm so sorry if i am bothering you)

"and i'm so sorry if i am bothering you" - no prob, nice to see you making progress.

you forgot the roi's now, try :

RL, Rr, PL, PR, Q, validRoiL, validRoiR = cv.stereoRectify(....

okay i really appreciate,now it works,thank you now moving to the next step thank you again

i'm getting this error: colors = cv.cvtColor(imgL, cv.COLOR_BGR2RGB) error: ......\src\opencv\modules\imgproc\src\color.cpp:3346: error: (-215) scn == 3 || scn == 4 :/ ?!!