I have become increasingly interested in artifical intelligence, but I am still learning the basics of the windows API. My current goal is to create a c++
process which calls SendInput
to simulate PrtSc being pressed (print screen button). This is the only way I currently can think of being able to have a program be able to see one frame of the screen as human might.
As a simple example, lets say, for some reason, I wanted my program to "look and see for itself," what time the clock on Windows 7 says. So to make the task easy for the computer and for myself, lets say we know the toolbar is on the bottom of the screen. We already know the exact screen coordinates of the rectangular bitmap we want to examine. So we call SendInput
to simulate pressing PrtSc
and copy the screen display to the clipboard and we examine the clipboard at those coordinates and do some feature-detection to see what numbers there are, and the rest is downhill from there.
So my question then, is this a reasonable way to implement such a program? Or am I making this more difficult than it needs to be?