OpenCV 3 Computer Vision with Python Cookbook

How it works

cv2.undistort removes lenses distortion from the image. It takes the following arguments: the image to undistort, the camera matrix, the distortion coefficients array, an object to store the undistorted image, and an optimal camera matrix. The last two arguments are optional. The function returns the undistorted image. If you missed the last parameter of cv2.undistort, the resulting image will contain empty (of black color) regions. The optimal camera matrix argument allows you to get the image without these artifacts, but we need a way to compute this optimal camera matrix, and OpenCV serves it.

cv2.getOptimalNewCameraMatrix creates an optimal camera matrix to get rid of black regions on the undistorted image. It takes the camera ...