Closed MugleLei closed 4 years ago
This method isn't taking into account lens distortion when stitching the images. This is okay on mobile phone footage, but it looks like you're using a lens with distortion.
Usually you have to undistort the images and stitch them.
video_stitching has a bug: Traceback (most recent call last): File "video_stitching.py", line 46, in
cap = cv2.VideoCapture(args.video_path)
TypeError: an integer is required
so I use image_stitching to stitch images captured from video, but I can't stitch 35 images at a time while stitch 34 images are fine. Also when stitching 80 images, No.1-34 disappear and No.35-80 are fine. I run the program in both opencv2.4 and opencv3.2 and they have the same problem. Is there any limitation in the code? I'm not familiar with opencv so I can't find the limitation myself.
when stitching 34 images:
when stitching 35 images:
when stitching 36 images:
when stitching 80 images: