Recognition of objects in the image and checking them for compliance with the "sample"

Question

C++ / C#

Hello!
There is a composite image, in it the text and a photo. There are about a hundred such images, they differ in different sizes and positioning of elements.
I know what the text should be and I have a photo to use.

It is clear that you need to find an image, text and compare them with samples.
And then a bunch of stupid questions: how to find; what to do with different sizes?
What approaches can be taken to solve these problems? Neural networks, state machine?
What literature to read? Preferably Russian ;)
Also, advise which database to use, I'm thinking of writing in C #
_{I think I'm asking such stupidity, I'm already ashamed.}

Answer 1

There is a good Open CV library and wrappers for C# -
Emgu CV , opencvdotnet , opencvsharp