Lela Star
Oct 25, 20201 min
For some tasks like tagging photos, scanning x-rays for signs of cancer, and detecting a braking car on the road in front of you, computer vision software is on par with the human vision system.
This is all the more impressive when you consider that computers “see” only by extracting patterns from arrays of numbers called pixels.
When light enters a computer’s camera from an autumnal scene like this:
it hits sensors that measure the incoming light’s intensity, and the result is a representation of the image that looks more like this:
… a slight less scenic.
Where we start:
Here’s a small region of a photograph (~1% of its pixels) that's been blown up so we can see the individual pixels.
Can you figure out what it is? Submit your guess below, and see if you are right!