Ask Your Question

How to avoid potential noise after Canny?

asked 2020-12-09 01:20:19 -0500

AndrewM gravatar image

updated 2020-12-09 09:13:22 -0500

I have a time averaged grayscale frame of a video: image description

I applied Canny edge detector to that frame, but sometimes there is noise like this: image description

I want to extract the area of a tv logo using boundingRect(), but it works wrong if there is noise. How can i avoid it?

edit retag flag offensive close merge delete


if you share your code maybe we can help to improve

sturkmen gravatar imagesturkmen ( 2020-12-09 03:23:40 -0500 )edit

1 answer

Sort by ยป oldest newest most voted

answered 2020-12-09 11:15:18 -0500

crackwitz gravatar image

simple average isn't good enough. you need to be sensitive to a pixel changing at all (within some tolerance) over a certain time, because the station logo doesn't change at all, while the rest of the picture usually changes, but may not change much, or may be visible for a long time but not all the time (that recognizable frame/border within view)

look into the background subtraction/segmentation module within OpenCV. tweak the parameters.

if that doesn't help, you'll probably need to come up with your own approach.

edit flag offensive delete link more

Question Tools

1 follower


Asked: 2020-12-09 01:20:19 -0500

Seen: 150 times

Last updated: Dec 09 '20