CODEWITHSUNDEEP
algorithmopencvimage-processingcomparisongoogle-image-search
Gosling
Gosling

Reputation: 41

What's the most accurate algorithm in finding the similarities between different images

I'm learning to find a way to search similar images recently. There is some popular algorithms in features matching area.For example, Perceptual Hash Algorithm, SIFI and SURF in openCV. I'm wondering that which one is the most accurate.Or is using multiple algorithms a good idea? Or is there some nice conclusions about the popular algorithms. Thanks in advance.

Upvotes: 2

Views: 467

Answers (1)

HMann
HMann

Reputation: 42

There are a lot of algorithms for check similarities, actually matching features. I searched some algorithms to find the features which are SURF, SIFT, BRISK, LBP, Harris MSER, A-KAZE, FAST and so on.

In many applications, the SIFT is selected to check feature matching. However, I think that you should evaluate the performances algorithm. To find right algorithm for your application.

If you can't evluate the algorithms, I think that using multiple algorithms is better to you.

If you want to check the features, I recommend this link to understand feature extractors, descriptors, matching. https://kr.mathworks.com/help/vision/local-feature-extraction.html

Thank you.

Upvotes: 2

Related Questions