python - Python:solvePnP()没有足够的值可解包？

Question

我在使用OpenCV中的cv2.solvePnP函数时遇到问题。此功能用于获取棋盘的姿势估计。在以下代码之后，我得到一个错误:

for fname in glob.glob('Images/Calibragem/img1*.jpg'):
    img = cv2.imread(fname)
    gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
    ret, corners = cv2.findChessboardCorners(gray, (9,6), None)

    if ret==True:

        corners2=cv2.cornerSubPix(gray,corners,(11,11),(-1,-1), criteria)

        #finds the vectors of rotation and translation
        ret, rotationVectors, translationVectors, inliers = 
            cv2.solvePnP(objp, corners2, matrix, distortion)
        #projects the 3D points in the image

        imgpts,jac = cv2.projectPoints(axis,rotationVectors,translationVectors,matrix,distortion)

        imgAxis=drawAxis(img,corners2,imgpts)
        cv2.imshow('imgAxis', imgAxis)
        cv2.imwrite('imgAxis.png',imgAxis)

错误提示:

ret, rotationVectors, translationVectors, inliers = cv2.solvePnP(objp, corners2, matrix, distortion) ValueError: not enough values to unpack (expected 4, got 3)

Answer 1

从opencv2 documentation:

 Python: cv2.solvePnP(objectPoints, imagePoints, cameraMatrix, distCoeffs[, rvec[, tvec[, useExtrinsicGuess[, flags]]]]) → retval, rvec, tvec¶

因此，只有3个值需要解压。
因此，您应该能够解决:

ret, rotationVectors, translationVectors = 
            cv2.solvePnP(objp, corners2, matrix, distortion)

由于solvePnP()仅返回 retval， rvec和 tvec。