1
$\begingroup$

I would like to compare 2 generic images, with same sizes and normalized values. Which metric would be better than the baseline Eucledian distance?

$\endgroup$
2
  • 1
    $\begingroup$ I understand that your images are not related. So each metric uses only one image, so this is not a distance. Then how are they normalized, and what do you want to use the metric for? $\endgroup$ – Laurent Duval Dec 1 '20 at 22:31
  • $\begingroup$ I want to determine how similar the images are. In the broader sense, I have image A1 and then compare it with images A2 to An and then sum all the similarity values. This can be done very easily using an Euclidean metric. $\endgroup$ – desert_ranger Dec 1 '20 at 22:35
3
$\begingroup$

One of the most known image similarity metric is the SSIM. You can take a look to the next links:

EDIT 1: Okay, you refer to semantic similarity. I have not worked on that issue really, and probably there are new DNN solutions to that purpose. I have not checked the state-of-the-art related to this topic. Even so, I have made a quick test using the skimage SSIM and the difference between images to give you some idea about what happens with these metrics on several images. Find in the next images some hints about their performance for your target application:

enter image description here

enter image description here

Please, notice the values of the SSIM, MSE and STD DIFF on the images. SSIM and MSE works very well to compare quality of the same image. But not for semantic comparison.

Cheers.

$\endgroup$
2
  • $\begingroup$ How good is SSIM for different images? My understanding is that SSIM is used for comparing the quality of 2 images. $\endgroup$ – desert_ranger Dec 1 '20 at 20:16
  • $\begingroup$ @desert_ranger I have edited my post to answer to your new question here, and clarify that you referred to semantic image comparison. I hope those experiments help you or give you some other ideas. $\endgroup$ – Arritmic Dec 1 '20 at 22:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.