I want to build an iOS application that recognizes patterns in a cup of coffee.
For example http://en.wikipedia.org/wiki/File:Coffereading.jpg .
This image recognition script could run on the client side (iphone eventually) or on server side.
The goal of the app is to take a picture of a cup of coffee, analyse patterns,
and compare them with images/patterns that are already stored in a database, and return the most appropriate one
I have no experience in this field and after doing some research, I found some libraries that might help me do this: openCV, kooaba, snaptell, and server side libraries like afroge.net.
This confused me alot. I want to know if this is possible using libraries like the ones above, or any other one. And how much time/effort should be given if any modification should be made on one of those in order to achieve my goal
You can also read more information about reading fortune under this link: http://en.wikipedia.org/wiki/Tasseography
In short: It can be done with OpenCV. You can use it to recognize the coffee cup (circle and/or square detection). After that you have to take a look at “feature detection” (SURF ist a good way to go) to match the coffee patterns.