In brief, for those who don't know Diablo 2, it is an action role-playing hack-and-slash computer video game.
The functions of the program should include, to detect paths, move the character around, click to items in the game, and use keys on keyboard when suitable, and so on.
I plan to use computer vision algorithms to detect objects on the game, and then set rules to perform suitable actions. How can I obtain what is on the screen? what libraries and subjects I should look for?