The supermassive black hole M87 is the first to have a photo of its own. Now, thanks to artificial intelligence (AI), its quality has been boosted to full resolution.
Full-Resolution Image of Supermassive Black Hole M87
According to Live Science, the M87's iconic shot was taken by pooling the radio light that moved 53 million light years across space. The shot captures the black hole, which is as big as a solar system, at the Virgo galaxy cluster's core.
Now, efforts have focused on cleaning and boosting the image with machine learning. It has been sharpened to achieve the fullest resolution possible. By doing so, a darker and bigger center has been seen to be surrounded by glowing gas. MailOnline notes that it has now been described as a "skinny donut."
The updated image was published in the Astrophysical Journal Letters.
1st image of black hole gets makeover thanks to Artificial Intelligence https://t.co/FN3TFNo39w
— ABC7 Eyewitness News (@ABC7) April 13, 2023
According to Sky News, the resolution of the boosted picture was enhanced via "image reconstruction algorithms" to help bridge the gaps that were left in the first observations.
Lia Medeiros, the study's lead author and an astronomer from the Institute for Advanced Study in Princeton, says that with PRIMO, their new machine learning method, they were able to achieve the highest possible resolution of the current array. Because they are incapable of closely studying black holes, the detailed shot has a significant role in helping them know more about the objects' behavior.
In the full-resolution shot, the ring's width is smaller by a factor of two. Medeiros adds that this will serve as a strong constraint for their theoretical models and gravity tests.
How AI Sharpened the Image of a Supermassive Black Hole
M87, or the Messier 87 black hole, has the width of the earth's solar system and the mass of 6.5 billion suns. As per Live Science, the original shot was imaged by the EHT (Event Horizon Telescope), which is an array of eight radio telescopes that are globally synchronized.
To sharpen the image, the researchers used principal-component interferometric modeling (PRIMO), which is an AI technique that analyzed over 30,000 high-fidelity simulated photos of gas accretions around black holes to spot patterns that are common.
Such patterns were then organized based on their frequency of occurrence before being mixed and applied to the original photo to sharpen it. As the researchers examined the rendered photo with theories regarding black hole appearance and EHT data, they concluded that the boosted image closely approximated the actual supermassive black hole, as noted by Live Science.
This newly enhanced shot will foster an even more in-depth study of these black holes and their effects, as they are where theories of quantum mechanics and gravity break down and get mixed with each other.
Medeiros notes that the 2019 shot of M87 is just the start. The PRIMO will serve as a vital tool in learning more about the stories behind such images.
RELATED ARTICLE: Second-Closest Supermassive Black Hole May Be Lurking in a Nearby Dwarf Galaxy About 1 Million Light-Years Away
Check out more news and information on Space in Science Times.