This year, my team would like to use more advanced vision processing than previous years. Which program is the most useful? Also, is either of them able to compute how the robot should move?
hmmmm ill tell you my experiences with all of them and let you decide:
Grip: Interesting program, very easy to use gui and good selection of algorithms, unfortunately in alpha and is still in development which means some parts are not working properly. Example: I have a uvc webcam that will work with opencv and mjpeg streamer but wont actually work with grip.
Opencv: Plenty of documentation out there for this library, is what Grip is based on, and runs on multiple languages (c++,python,java,C).
RoboRealm: While as cool as it is to have a program do the work for you, think about this: Will you be able to put it on your resume? Will the judges think its orriginal? are you ok with not being able to optimize it?
We will not be using that advanced features of a standard computer vision library so its really up to you…