The problem was in python bindings in void matchImagePoints(InputArrayOfArrays detectedCorners, InputArray detectedIds, OutputArray objPoints, OutputArray imgPoints).

detectedCorners may include aruco marker corners as vector<vector<Point2f>> or vector<Mat>. Any Mat in vector<Mat> must have 4 rows for 4 corners for one marker (one marker has 4 corners).

detectedCorners may include Charuco corners as vector<Point2f> or Mat. Python bindings add extra dimension to detectedCorners and therefore vector<Mat> case is additionally processed for charuco corners.

Until the PR merge, you can using custom matchImagePoints():

def matchImagePoints(charuco_board, charuco_corners, charuco_ids):
    objPoints = []
    imgPoints = []
    for i in range(0, len(charuco_ids)):
        index = charuco_ids[i]
        objPoints.append(charuco_board.getChessboardCorners()[index])
        # objPoints[-1][0][1] = charuco_board.getRightBottomCorner()[1] - objPoints[-1][0][1]  #  set old axis direction
        imgPoints.append(charuco_corners[i])
    return np.array(objPoints), np.array(imgPoints)

Use sample from #23139 to see problem.

args:
-w=5 -h=7 -sl=0.04 -ml=0.02 -d=10 -v=choriginal.jpg

With fix:

detectedCorners includes charuco corners in this case. Python bindings add extra dimension to detectedCorners and therefore without fix detectedCorners processed as aruco marker corners and matchImagePoints returns bad values:

Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

• I agree to contribute to the project under Apache 2 License.
• To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
• The PR is proposed to the proper branch
• There is a reference to the original bug report and related work
• There is accuracy test, performance test and test data in opencv_extra repository, if applicable
  Patch to opencv_extra has the same branch name.
• The feature is well documented and sample code can be built with the project CMake