Journal: Light, Science & Applications

Article Title: All-optical image denoising using a diffractive visual processor

doi: 10.1038/s41377-024-01385-6

Figure Lengend Snippet: Experimental setup for a 3-layer diffractive image denoiser. a Photograph of the experimental setup including the 3D-fabricated all-optical image denoiser trained for noisy intensity images. b Intensity profiles of an image example impacted by various levels of salt-only noise ( P te ) and their photographs after their 3D-fabrication. c Phase profiles of the trained diffractive image denoiser layers and their photographs after 3D-fabrication. d Schematic of the experimental setup using continuous wave THz illumination (λ = ~0.75 mm)

Article Snippet: The thickness profiles of the trained diffractive surfaces and noisy/clean input objects were converted into STL files using MATLAB and they were fabricated by using a 3D printer (Objet30 Pro, Stratasys Ltd.).

Techniques:

Journal: Light, Science & Applications

Article Title: All-optical image denoising using a diffractive visual processor

doi: 10.1038/s41377-024-01385-6

Figure Lengend Snippet: Experimental results of the all-optical diffractive image denoiser. a Layout of the diffractive image denoiser with 3 transmissive layers. b Photographs of 3D-fabricated layers of the trained diffractive image denoiser. c Experimental and numerical image denoising performance of the designed diffractive denoiser under different levels of salt-only noise ( P te ). The PSNR value for each case is shown beneath the respective image. Experimental results were normalized before the PSNR calculation

Article Snippet: The thickness profiles of the trained diffractive surfaces and noisy/clean input objects were converted into STL files using MATLAB and they were fabricated by using a 3D printer (Objet30 Pro, Stratasys Ltd.).

Techniques: