On April 10, 2019, a never-before-seen image of black hole M87 Galaxy has revealed thanks in significant part to the Total Variation (TV) Theory of Image Processing first proposed in the 1987 CALTECH PH.D. Thesis of Cognitech’s co-founder and CEO Dr. Leonid ( Lenny) Rudin.
Specifically the TV-based Image Denoising and Restoration Algorithm published in 1992 by Leonid Rudin, Stan Osher, Emad Fatemi, all members of Cognitech scientific team at the time, was a key factor in the success to “achieve an optimal resolution of ∼20%–30% of the diffraction limit λ/Dmax, which is the nominal spatial resolution of a radio interferometer”, accordingly to the Astrophysical Journal article by K. Akiyama Et. Al. (Event Horizon Telescope EHT team) as well as the announcement by the National Science Foundation ( NSF-IPAM): “Rudin-Osher-Fatemi Model Captures Infinity and Beyond”.
Cognitech’s Total Variation Algorithm
Total Variation Denoising and Restoration enables the accurate reconstruction of extremely noisy and blurry images, with specific accuracy in reconstructing sharp image features. With the help of the Total Variation regularization algorithm, the EHT was able to “robustly and reasonably achieve super-resolution sufficient to clearly resolve the black hole shadow.”, i.e. reveal the image of the black hole for the first time.
Total Variation based image processing algorithms are not just limited to astrophysics applications. Proprietary TV-based software tools have been designed by Cognitech’s scientists and engineers to be successfully applied to restore noisy, blurry, and resolution degraded CCTV videos arising in forensic and police investigation applications.
We hope you enjoyed this CogniKnowledge Blog! Stay tuned and don’t miss the next blogs. You can also follow us on Twitter, Facebook, Instagram, LinkedIn, or YouTube: we post CogniKnowledge Blogs and ForensicFriday Blogs regularly so you won’t miss any!