Recognize duplicate images in a larger

Question

Edit: this is not a duplicate. Determine if an image exists in a large image, and if so, find it using Python , since I do not know the template in advance

Suppose I have a large image (usually a photograph taken with a camera, so it may be a little noisy, but let it still not be accepted), consisting of several smaller images, all equal to each other, something like

I need to find the outline of each of them. The first step is to recognize a repeating image (or an unknown pattern) in the 2D image. How can I achieve this first step?

I read that I can use the FFT of the original image and look for repetitive frequencies, would this be an acceptable approach?

: , , . , . .

Answer 1

(Sift/SURF/ORB/etc.), . opencv, 3x , .

, .

, , "" . , , ( ), , . - . , " ", , / (. ). , , .

, :

< > :

• - , .

• (, , )
• - .

, , ;)

, , , .

Answer 2

blob ROI, . . SAD, SSD .

,

• /

• ...

; , "".

Answer 3

, , . .

• (, )?
• ? , .
• ( ) ? ?
• , , ?

Yves Daoust , , .

Micka , , .

: .

:

• MxN M, N ( , ).
• . . .
• ( ) .
• , ( ) (, ).
• , M, N .
• To speed up the launch of the entire process on a coarse scale (a smaller version of the image), and then process small scales only where necessary. Needs balancing scaling and performing correlations.

Sorry, I cannot do this full Matlab project right now, but I hope this helps you.