-
-
Notifications
You must be signed in to change notification settings - Fork 56.5k
drawMatches lines not visible in larger images #20095
Copy link
Copy link
Closed
Milestone
Description
drawMatches https://docs.opencv.org/master/d4/d5d/group__features2d__draw.html#gad8f463ccaf0dc6f61083abd8717c261a draws 1 pixel thick lines between matching keypoints in two side-by-side images, this becomes invisible for larger images.
Please add a thickness argument, and base the default thickness on the size of the image.
System information (version)
- OpenCV => 4.2
Steps to reproduce
```.py
left_image = cv.imread('images/IMG_6769.jpg',cv.IMREAD_COLOR)
right_image = cv.imread('images/IMG_6770.jpg',cv.IMREAD_COLOR)
width = right_image.shape[1]
height = right_image.shape[0]
orb = cv.ORB_create(nfeatures=1500)
kp1, des1 = orb.detectAndCompute(left_image,None)
kp2, des2 = orb.detectAndCompute(right_image,None)
bf = cv.BFMatcher(cv.NORM_HAMMING, crossCheck=True)
matches = bf.match(des1,des2)
bestmatches_diff_to_median_disparity_wise = get_best_matches_difference_to_median_disparity_wise(matches, kp1, kp2)
largest_trapeze_matches = get_4_matches_that_form_largest_trapeze(bestmatches_diff_to_median_disparity_wise, kp1)
destination_left_points, source_right_points = from_matches_to_2_arrays_of_points(largest_trapeze_matches, kp1, kp2)
img3 = cv.drawMatches(left_image,kp1,right_image,kp2,largest_trapeze_matches,None,flags=cv.DrawMatchesFlags_DRAW_RICH_KEYPOINTS) # cv.DrawMatchesFlags_NOT_DRAW_SINGLE_POINTS
plt.imshow(img3),plt.show()
```
Issue submission checklist
Reactions are currently unavailable

