I have hundreds of images I want to pull information out of. Each image contains a series (4 groups, aligned horizontally) of little icons that I want to translate into a text file. The program will know how to translate these icons into text via a simple definition I set up before, telling it what each icon means. The icons in the source images that I want to scan will always have the same perspective, size, coloring. So it should be a simple matching solution without involving some kind of AI.
Here are some pictures to visualize what I mean:
Example of image I want to scan and translate into text via a definition of what each icon means
Definition of what each icon means
So after setting up the definition in the program and having it scan the example image, the output should be something like this: Scopesmith Necromancer | Greyhound Necromancer Beastface | Fanning Determination Ghoul Bloodless | Fanning Levering Poacher Greyhound
Every google search for how to approach this leads me to AI, but I'm thinking there should be a less complex solution since it's simple image matching by a predefined set of rules. If someone could point me towards a tool for this problem that's already existing or towards what I should be searching for on google, it would help me out a lot. I have no experience in image processing but some experience in coding, so if I know how to describe the problem to a search engine I think I can manage.