Mastering OpenCV 4 with Python

Camera pose estimation and basic augmentation

At this point, we can overlay some images, shapes, or 3D models in order to see a complete augmented reality application. In this first example, we are going to overlay a rectangle with the size of the marker. The code to perform this functionality is the following:

if ids is not None:    # rvecs and tvecs are the rotation and translation vectors respectively    rvecs, tvecs, _ = cv2.aruco.estimatePoseSingleMarkers(corners, 1, cameraMatrix, distCoeffs)    for rvec, tvec in zip(rvecs, tvecs):        # Define the points where you want the image to be overlaid (remember: marker coordinate system):        desired_points = np.float32(            [[-1 / 2, 1 / 2, 0], [1 / 2, 1 / 2, 0], [1 / 2, -1 / 2, 0], [-1 / 2, -1 / 2, 0]])        # Project ...