Enrollments Open for Advanced Certification Courses on Data Science, ML & AI by E&ICT Academy IIT Roorkee

  Apply Now

Finding Homography Matrix

It’s time to align the images now.

We need to align the images because, though both images could form a panorama, the two could differ in terms of angle, translation, size, etc, probably caused due to orientational difference in the camera while capturing the photos. Thus, we need to find a transformation matrix to perform this alignment, which ensures compatibility to stitch and form a panorama.

A Homography Matrix is a 3x3 transformation matrix that maps the points in one image to the corresponding points in the other image.

enter image description here

where (x1, y1) and (x2, y2) could be thought of the co-ordinates of the matching key-points in both the images respectively.

A homography matrix is used with the best matching points, to estimate a relative orientation transformation within the two images. Simply put, it is the perspective transformation matrix between two planes - here, the two images.

In order to find a homography matrix, at least 4 matching key-points are needed.

enter image description here


cv2.findHomography finds a perspective transformation between two planes - here, the two images. More here.

cv2.RANSAC is a method used to compute a homography matrix.

  • Get the best matches - matches - between the right and left images. If there are at least 4 matches, find the Homography matrix. Else, raise an error.

    if len(matches) >= 4:
        src = np.float32([ kp1[m.queryIdx].pt for m in matches ]).reshape(-1,1,2)
        dst = np.float32([ kp2[m.trainIdx].pt for m in matches ]).reshape(-1,1,2)
        H, masked = cv2.findHomography(src, dst, cv2.RANSAC, 5.0)
        raise AssertionError("Can't find enough keypoints.")

No hints are availble for this assesment

Answer is not availble for this assesment

Note - Having trouble with the assessment engine? Follow the steps listed here

Loading comments...