如何将点从未变形图像投影到变形图像？

Question

[我借助cv :: fisheye :: calibrate使鱼眼镜头图像失真，并发现以下系数。

K = 
array([[541.11407173,   0.        , 659.87320043],
       [  0.        , 541.28079025, 318.68920531],
       [  0.        ,   0.        ,   1.        ]])
D =
array([[-3.91414244e-02],
       [-4.60198728e-03],
       [-3.02912651e-04],
       [ 2.83586453e-05]])

new_K = cv2.fisheye.estimateNewCameraMatrixForUndistortRectify(K, D, (1280, 720), np.eye(3), balance=1, new_size=(3400, 1912), fov_scale=1)
map1, map2 = cv2.fisheye.initUndistortRectifyMap(K, D, np.eye(3), new_K, (3400, 1912), cv2.CV_16SC2)
undistorted_img = cv2.remap(distorted_img, map1, map2, interpolation=cv2.INTER_LINEAR, borderMode=cv2.BORDER_CONSTANT)

如何找到x和y？

Answer 1

objp = np.array([[[(1595-new_K[0, 2])/new_K[0, 0], (922-new_K[1, 2])/new_K[1, 1], 0.]]])
rvec = np.array([[[0., 0., 0.]]])
tvec = np.array([[[0., 0., 0.]]])
imgpoints2, _ = cv2.fisheye.projectPoints(objp, rvec, tvec, K, D)

x = imgpoints2[0,0,0]
y = imgpoints2[0,0,1]