Shulou(Shulou.com)11/24 Report--

CTOnews.com, April 13 (Xinhua) Black holes are gravitationally strong celestial bodies. They are so massive but so small that even light cannot escape. Scientists believe that the center of almost every large galaxy has a supermassive black hole with the mass of hundreds of thousands to billions of suns. Supermassive black holes release huge amounts of energy by gobbling up the surrounding gas, forming active galactic nuclei and quasars.

In 2019, the event Horizon Telescope (EHT) collaboration released the first direct photograph of a black hole in human history. This photo shows a supermassive black hole at the center of the galaxy M87, which is 5500 light-years from Earth and has a mass equivalent to 6.5 billion suns. What we see is a faint orange ring made up of radio waves emitted by gases heated to extremely high temperatures and spinning at close to the speed of light. The black area in the middle of the ring is the event horizon of the black hole and its shadow. Now the halo is clearer, from a fuzzy doughnut to a slender halo, thanks to the use of a machine learning tool called PRIMO.

The blurred image of the black hole does not show the highest resolution of the EHT telescope array because it is a splicing of data collected by seven telescopes in different locations around the world. To solve this problem, a research team used a new machine learning technique, the Principal component interference Imaging Model (PRIMO), to "fill in the gaps in the M87 image" and to maximize the resolution of the EHT telescope array for the first time.

According to CTOnews.com, PRIMO technology is realized through dictionary learning (Dictionary Learning), which allows computers to generate rules based on a large number of training materials. The team inputted 30,000 high-fidelity simulated images of black holes into PRIMO, covering a variety of theoretical predictions of how black holes accrete matter. PRIMO identifies black holes by looking for patterns and sorts them according to how often they appear in the simulation. These modes can then be combined with EHT images to generate a higher-resolution image of the M87 black hole and reveal structures that may be missing from the telescope array.

This PRIMO-processed image is consistent with EHT data and the theoretical black hole model. These models explain that the bright ring we see is because the gas is accelerated to close to the speed of light by the strong gravity of the black hole, and is therefore heated and glowed, rotating near the event horizon, the boundary where the black hole cannot escape light.

"after EHT released the first image of a black hole on the event horizon scale in 2019, we reached another milestone, producing an image that used the full resolution of the array for the first time," said Psaltis, a researcher on the project. "the new machine learning technology we developed provides a golden opportunity for our collective work to understand black hole physics."

Using PRIMO to improve the resolution of EHT images could help better estimate the characteristics of the two supermassive black holes, including their mass, size and the speed at which they engulf matter, team members said. New machine learning techniques can also be applied to other observations, such as EHT's image of Sagittarius A, the supermassive black hole at the center of our Milky way galaxy, in 2022.

The new study was published online today (April 13) in the Astrophysical Journal Letters.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.