Python Image Processing Cookbook

Panorama with OpenCV-Python

You can implement image-stitching directly with the OpenCV-Python library's Stitcher class methods (without explicitly performing the stitching steps, for example, feature extraction/matching/blending). Just run the following steps:

1. First, ensure that the version of OpenCV you are using is <= 3.4.2 (until the SIFT patent expires):

print(cv2.__version__)# 3.4.2

2. Initialize the list of the images to be stitched:

images = [ cv2.cvtColor(cv2.imread(img), cv2.COLOR_BGR2RGB) for \                    img in glob.glob('images/victoria*.png')]print('Number of images to stitch: {}'.format(len(images)))# Number of images to stitch: 7

3. Instantiate OpenCV-Python library's image stitcher object and then use the stitch() method to generate the ...