I know Google Image Search is very powerful by their face recognition algorithm.
I wonder is there any way to process/change the images which Google cannot search or detect?
I tried to change the images (of some famous movies like Batman) to black&white, reduce the resolution or "print-screen" each scene from the movies but they didn't work. Google can find out all of them.
I assume that Google will act the same for text and image crawling. The easiest way should be to define a robots.txt for your site.
This file can look like this:
User-agent: *
Disallow: /images/
By setting this up, google will not index everything in and below /images. Besides google every other well-behaving search engine crawler will consider your robots.txt.
See details about this file here.