Ask Your Question

Blur detection for focused images

asked 2017-06-19 04:01:58 -0500

I am finding out if the images are blur or not using OpenCV Laplacian method. For normal images, it is working fine. But when I am using object focused images, it is identifying as Blur image. So is there any method that I can apply for object focused images? Sample image with low blurry value

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2017-06-19 05:10:48 -0500

pklab gravatar image

updated 2017-06-19 05:35:39 -0500

Blur with Laplacian method returns and indirect measure of Blur. It's relative to the image context therefore you can't set absolute bias.

If you compare values from same subject one focused other out of focus you can get best focused using Laplacian method.

BTW your image contains large blurred area (background) that reduces the "Laplacian focus score".

  • You might preprocess the image to separate the background from foreground than evaluate the "Laplacian focus score" only on foreground.
  • As 2nd way you might calculate local "Laplacian focus score" maybe split the image in 3x3 sub-images, calculate your score for each sub-image. Maybe central sub-images are the foreground and give you better results.
edit flag offensive delete link more

Question Tools

1 follower


Asked: 2017-06-19 04:00:10 -0500

Seen: 1,459 times

Last updated: Jun 19 '17