MSGID: 1:317/3 6427b37f   
   PID: hpt/lnx 1.9.0-cur 2019-01-08   
   TID: hpt/lnx 1.9.0-cur 2019-01-08   
    AI algorithm unblurs the cosmos    
    Tool produces faster, more realistic images than current methods    
      
     Date:   
         March 31, 2023   
     Source:   
         Northwestern University   
     Summary:   
         Researchers adapted a well-known computer-vision algorithm   
         used for sharpening photos and, for the first time, applied   
         it to astronomical images from ground-based telescopes. While   
         astrophysicists already use technologies to remove blur, the adapted   
         AI-driven algorithm works faster and produces more realistic images   
         than current technologies. The resulting images are blur-free and   
         truer to life.   
      
      
         Facebook Twitter Pinterest LinkedIN Email   
   FULL STORY   
   ==========================================================================   
   The cosmos would look a lot better if Earth's atmosphere wasn't photo   
   bombing it all the time.   
      
      
   ==========================================================================   
   Even images obtained by the world's best ground-based telescopes are   
   blurry due to the atmosphere's shifting pockets of air. While seemingly   
   harmless, this blur obscures the shapes of objects in astronomical images,   
   sometimes leading to error-filled physical measurements that are essential   
   for understanding the nature of our universe.   
      
   Now researchers at Northwestern University and Tsinghua University in   
   Beijing have unveiled a new strategy to fix this issue. The team adapted   
   a well-known computer-vision algorithm used for sharpening photos and,   
   for the first time, applied it to astronomical images from ground-based   
   telescopes. The researchers also trained the artificial intelligence (AI)   
   algorithm on data simulated to match the Vera C. Rubin Observatory's   
   imaging parameters, so, when the observatory opens next year, the tool   
   will be instantly compatible.   
      
   While astrophysicists already use technologies to remove blur, the   
   adapted AI- driven algorithm works faster and produces more realistic   
   images than current technologies. The resulting images are blur-free   
   and truer to life. They also are beautiful -- although that's not the   
   technology's purpose.   
      
   "Photography's goal is often to get a pretty, nice-looking image,"   
   said Northwestern's Emma Alexander, the study's senior author. "But   
   astronomical images are used for science. By cleaning up images in   
   the right way, we can get more accurate data. The algorithm removes   
   the atmosphere computationally, enabling physicists to obtain better   
   scientific measurements. At the end of the day, the images do look better   
   as well."  The research will be published March 30 in the Monthly Notices   
   of the Royal Astronomical Society.   
      
   Alexander is an assistant professor of computer science at Northwestern's   
   McCormick School of Engineering, where she runs the Bio Inspired   
   Vision Lab.   
      
   She co-led the new study with Tianao Li, an undergraduate in electrical   
   engineering at Tsinghua University and a research intern in Alexander's   
   lab.   
      
   When light emanates from distant stars, planets and galaxies, it travels   
   through Earth's atmosphere before it hits our eyes. Not only does our   
   atmosphere block out certain wavelengths of light, it also distorts the   
   light that reaches Earth. Even clear night skies still contain moving   
   air that affects light passing through it. That's why stars twinkle and   
   why the best ground-based telescopes are located at high altitudes where   
   the atmosphere is thinnest.   
      
   "It's a bit like looking up from the bottom of a swimming pool," Alexander   
   said. "The water pushes light around and distorts it. The atmosphere is,   
   of course, much less dense, but it's a similar concept."  The blur becomes   
   an issue when astrophysicists analyze images to extract cosmological   
   data. By studying the apparent shapes of galaxies, scientists can detect   
   the gravitational effects of large-scale cosmological structures, which   
   bend light on its way to our planet. This can cause an elliptical galaxy   
   to appear rounder or more stretched than it really is. But atmospheric   
   blur smears the image in a way that warps the galaxy shape. Removing   
   the blur enables scientists to collect accurate shape data.   
      
   "Slight differences in shape can tell us about gravity in the universe,"   
   Alexander said. "These differences are already difficult to detect. If   
   you look at an image from a ground-based telescope, a shape might   
   be warped. It's hard to know if that's because of a gravitational   
   effect or the atmosphere."  To tackle this challenge, Alexander and   
   Li combined an optimization algorithm with a deep-learning network   
   trained on astronomical images. Among the training images, the team   
   included simulated data that matches the Rubin Observatory's expected   
   imaging parameters. The resulting tool produced images with 38.6% less   
   error compared to classic methods for removing blur and 7.4% less error   
   compared to modern methods.   
      
   When the Rubin Observatory officially opens next year, its telescopes   
   will begin a decade-long deep survey across an enormous portion of the   
   night sky.   
      
   Because the researchers trained the new tool on data specifically   
   designed to simulate Rubin's upcoming images, it will be able to help   
   analyze the survey's highly anticipated data.   
      
   For astronomers interested in using the tool, the open-source,   
   user-friendly code and accompanying tutorials are available online.   
      
   "Now we pass off this tool, putting it into the hands of astronomy   
   experts," Alexander said. "We think this could be a valuable resource   
   for sky surveys to obtain the most realistic data possible."  The study,   
   "Galaxy image deconvolution for weak gravitational lensing with unrolled   
   plug-and-play ADMM," used computational resources from the Computational   
   Photography Lab at Northwestern University.   
      
       * RELATED_TOPICS   
             o Space_&_Time   
                   # Galaxies # Space_Telescopes # Astronomy # Cosmic_Rays   
             o Computers_&_Math   
                   # Photography # Computers_and_Internet #   
                   Information_Technology # Hacking   
       * RELATED_TERMS   
             o Computer_vision o Quantum_computer o 3D_computer_graphics   
             o Fractal o Computer-generated_imagery o Webcam o   
             Computer_animation o MRAM   
      
   ==========================================================================   
   Story Source: Materials provided by Northwestern_University. Original   
   written by Amanda Morris. Note: Content may be edited for style and   
   length.   
      
      
   ==========================================================================   
   Journal Reference:   
      1. Tianao Li, Emma Alexander. Galaxy Image Deconvolution for Weak   
         Gravitational Lensing with Unrolled Plug-and-Play ADMM. Submitted   
         to arXiv, 2023 DOI: 10.48550/arXiv.2211.01567   
   ==========================================================================   
      
   Link to news story:   
   https://www.sciencedaily.com/releases/2023/03/230331120633.htm   
      
   --- up 1 year, 4 weeks, 4 days, 10 hours, 50 minutes   
    * Origin: -=> Castle Rock BBS <=- Now Husky HPT Powered! (1:317/3)   
   SEEN-BY: 15/0 106/201 114/705 123/120 153/7715 226/30 227/114 229/110   
   SEEN-BY: 229/111 112 113 307 317 400 426 428 470 664 700 292/854 298/25   
   SEEN-BY: 305/3 317/3 320/219 396/45   
   PATH: 317/3 229/426