He also ate a fresh Apple, so fresh orange is more like it.
Because dark and light is characterized in that in the dark can not see anything, not that dark hex number #000000---- they are just so different that after scanning the image for further processing, including programs that implement what is called AI color accept different numeric designations. And more - from the point of view of computer processing - they just and does not differ.
How to teach a machine to distinguish and understand what an Apple is--- so, "to distinguish" or "to understand." To distinguish - see something called "methods of recognition of visual images". Today - a purely technical, almost routine task. Dozens of implementations and hundreds of thousands of sources on the web, where everything is described in detail. "Understand" - and no AI really nothing to "understand". Don't believe - well, then give a definition of what you mean by "understand".
A simple example of a visual--- and what is there to understand. And why would the AI understand what "fresh"?
How to teach a machine to distinguish and understand what an Apple is, if it cannot eat?
Find more questions by tags Machine learning