This project is a little computer vision experiment I worked on in early 2019 The python file features two classes: the camera and the sensor The program uses your computer's webcam if available, and will crash if you do not have one The camera detects changes in its environment, and can mark target locations as items with the most movement. This is easily seen if moving your head while the program is running. The sensor is more accurate if the webcam is still
You can tinker with the drawn output on the screen by using drawOnScreen, drawVisionOnScreen, blitTarget and blitTargets available in the Sensor class
The program is not built for performance, and is not be suitable for real-time applications. It was only intended as an experiment
The camera records the world using the webcame, and detects changes between frames, which produces a monochrome image with white = high change, black = no change
The camera detects all pixels with a change above the specified threshold. For efficiency, a target is randomly chosen to be the "current target", but this could be improved
The camera's vision is drawn on screen, then the camera's targets are drawn over this to show the output
REQUIREMENTS:
opencv for python (python3 -m pip install opencv-python)
pygame (python3 -m pip install pygame)
@Author Harrison Boyle-Thomas @Date: ~2019