Best approach between Template Matching and Image Recognition [closed]
Hello guys.
I am trying to identify the heroes used in a mobile Android game at runtime. This game has thousands of heroes so I feel like Template matching (http://docs.opencv.org/master/de/da9/...) is not the most efficient way of dealing with this because I would have to find a match for each of them. Scaling might also be an issue depending on the device. I also took a look at Image Recognition but I don't understand how I could do this in my case (offline and using custom images)
Any hint to what to look for would be welcome. Thanks
sorry, but so far it's only an "I WANT" thing, - hot air.
convince us, that you DID some previous research, put in some effort to solve it, -- THEN we'going to help you.
So you mean I need to try an approach that might not be the best and then ask the same question again?
just read the comment, again.
Yea I did it many times trying to understand what I have to do to convince you :D I've been searching for days now, reading on many approaches (Tensorflow, OpenCV, SikuliX, ...) but still none of them seems to fit my situation.
which IS, in the 1st place ? do you realize, that noone can help you unless you:
I want to create an overlay app for Android that can add information about the fight based on the current leaders of the team. I need to identify the leaders by recognizing the icons that are on-screen at a specific spot. I know I have not started coding yet but I felt like I had to find a potential solution before heading in.
Just saw your edit. I get it man. I will get my hands dirty right now using a tutorial and trying to use it to solve my problem. Thanks!