There doesn't exist a way to infer the individual elements of a rendered frame in canvas because it just encodes the pixels. You can't know which pixels make up

There doesn't exist a way to infer the individual elements of a rendered frame in canvas because it just encodes the pixels. You can't know which pixels make up a text or button, that information is lost, unlike a DOM tree.