OpenCV 4 with Python Blueprints - Second Edition

Finding camera parameters

In order to be able to merge images, we need to compute homography matrices of all the images and then use those to adjust the images so they can be merged together. We will write a function to do that:

1. First, we are going to create the HomographyBasedEstimator() function:

def find_camera_parameters(features, pair_matches):    estimator = cv2.detail_HomographyBasedEstimator()

2. Once we have the estimator, for extracting all the camera parameters, we use the matched features from different images:

    success, cameras = estimator.apply(features, pair_matches, None)    if not success:        raise RuntimeError("Homography estimation failed.")

3. We make sure the R matrices have the correct type:

    for cam in cameras: cam.R = cam.R.astype(np.float32) ...