What specifically are the corners and sizes used by cv.detail.resultRoi()?

asked 2020-06-25 21:06:56 -0500

cmch gravatar image

updated 2020-06-26 03:25:41 -0500

supra56 gravatar image


I am looking at the stitching_detailed.py code provided by OpenCv for Stitching stitching_detailed.py. I am trying to understand how the blending works and I am having a hard time to understand what are the corners and sizes obtained after warping the image and the mask.
Specifically this line:

dst_size = cv.detail.resultRoi(corners=corners, sizes=sizes)

When running through the code to stitch 3 images for example I get a list of corners with one set for each image:

[(-3133, 64740), (-2300, 65756), (-1446, 64768)]

I looked at the OpenCV documentation and I did not get much insight. C:\fakepath\resultRoi().png

Is there another method of calculating those corners and sizes? What specifically are they? I am sorry for my misunderstanding, I am new to programming. Thank you!

edit retag flag offensive close merge delete