Autonomous Dino run using OpenCV and Numpy. The Running t-rex of 21st century.
..........................................................................
It works by capturing a frame in front of the dino and comparing it with a reference frame of totally white or black pixels. And if an obstacle is found, Spacebar is triggered and the Dino jumps. Also the Region of Interest(RoI) for frame capture is updated to adapt with the increasing speed.
The capturing Region of Interest must be tuned as per your display. By default it is tuned for 19" and chrome in half screen at right (as in figure)
The RoI updation could be tuned, to go beyond the High Score. It won't be hard to go till 10,000.