Here's how scientists will take the first-ever photograph of a black hole

You've probably seen a handful of amazing black hole images on the internet, like the one below.

It shows a black hole dragging in the hot plasma of a nearby blue giant star, never to be seen again:

Wikimedia Commons

Or perhaps you've seen this illustration, which shows the immense mass of a black hole warping the space around it into a gravitational lens:

Wikimedia Commons

And here's a view of a black hole belching particles near the speed of light — also called relativistic jet or "blazar":

Wikimedia Commons

No matter how awe-inspiring these images are, though, they're fantasy.

We have no idea how the event horizon (or boundary) of a black hole might actually look to the human eye, since the gravity of black holes is so strong that not even light can escape. Black holes are also incredibly far away, and usually very compact. And we don't really know what happens as an object approaches an event horizon, or what, if anything, might be on the other side.

But scientists are working to clear those hurdles — and soon — with the help of artificial intelligence.

The Massachusetts Institute of Technology (MIT) announced on Thursday that researchers have developed a learning computer algorithm that could help astronomers piece together the first-ever image of a black hole. They'll present the project on June 27, 2016, at the Conference on Computer Vision and Pattern Recognition in Las Vegas.

Astronomers have so far observed black holes only by the behavior of the objects around them. Stars, as you've probably noticed, emit plenty of visible light (as well as X-rays and radio waves), and we've developed advanced telescopes to measure black holes' effects on nearby stars.

Based on these observations, this is how scientists think a black hole may look as it warps the space around it:

Wikimedia Commons

But the new algorithm, developed by a collaboration between MIT and Harvard University could give us our first real, direct view.

The project will start with the international Event Horizon Telescope project, which "seeks, essentially, to turn the entire planet into a large radio telescope dish," MIT said in its announcement.

The project, called the Event Horizon Telescope, will gather data from a network of existing radio telescopes from all over the world — but because of the scale of the search, these images still won't be enough.

"A black hole is very, very far away and very compact," MIT researcher Katie Bouman said in the release. "It's equivalent of looking at a grapefruit on the moon, but with a radio telescope."

The team's algorithm will compare measurements from three telescopes, filtering out as much "noise" as possible in data that come from a giant black hole at the center of the Milky Way galaxy. The program will then analyze those radio signals to reconstruct a sharp image of the object.

Think of it like software that can complete a giant jigsaw puzzle with a bunch of missing pieces.

You may know the puzzle is supposed to show a face, but about half the pieces are gone. You assemble what you can with some precision, but there are still holes in the image. To finish the puzzle, you could create a computer program that knows more or less what faces look like, and understands that what it sees in the puzzle is almost certainly a face.

Bouman and her team's new program works similarly, reconstructing the "puzzle" of what a black hole looks like without having access to all of the data we'd like to.

Wikimedia Commons

To detect black holes today, computer-powered observatories scan for and record bright points of light that are emitted as a black hole, say, eats a star's plasma.

The new model will take such data about known black holes to identify common patterns among the enigmatic objects. Then the software will "learn" those patterns and use them to predict what appears in areas we can't see using radio telescopes.

If the project succeeds, it will create the first-ever direct image of a black hole. It's also expected to provide further insight as to whether Albert Einstein's general theory of relativity accurately predicts the behavior of black holes.

The team behind the project announced in January that it plans on taking the first picture ever of the (imaginary) surface of a black hole in 2017.

They'll be setting their sights on Sagittarius A: the giant black hole at the center of the Milky Way Galaxy with a mass of 4 million suns.