As described in https://docs.opencv.org/4.4.0/d4/d94/tutorial_camera_calibration.html: x_{distorted} = x( 1 + k_1 r^2 + k_2 r^4 + k_3 r^6) y_{distorted} = y( 1 + k_1 r^2 + k_2 r^4 + k_3 r^6) x_{distorted} = x + [ 2p_1xy + p_2(r^2+2x^2)] y_{distorted} = y + [ p_1(r^2+ 2y^2)+ 2p_2xy]
But in https://en.wikipedia.org/wiki/Distortion_(optics), Brown–Conrady model:
the above two models are in two different directions. I'm pretty confused when being told that OpenCV uses Brown–Conrady model for distortion.