OpenCV Module 9: Image Stitching and Creating Panoramas

Creating Panoramas using OpenCV

1. Find keypoints in all images

2. Find pairwise correspondences

3. Estimate pairwise Homographies

4. Refine Homographies

5. Stitch with Blending

We can create panoramas using the stitcher class

import cv2
import glob
import matplotlib.pyplot as plt
import math

#Read images
imagefiles = glob.glob("boat/*")
imagefiles.sort()

images = []
for filename ine imagefiles:
	img = cv2.imread(filename)
	img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
	images.append(img)

num_images = len(images)

#Display Images
plt.figure(figsize=[30,10])
num_cols = 3
num_rows = math.ceil(num_images / num_cols)
for i in range(0, num_images):
	plt.subplot(num_rows, num_cols, i+1)
	plt.axis('off')
	plt.imshow(images[i])

#Stitch Images
stitcher = cv2.Stitcher_create()
status, result = stitcher.stitch(images)
if status == 0:
	plt.figure(figsize=[30,10])
	plt.imshow(result)