Source-linked AI summary
High-index dielectric metasurfaces performing mathematical operations
Andrea Cordaro, Hoyeong Kwon, Dimitrios Sounas, A. Femius Koenderink, Andrea Alù, Albert Polman
TL;DR
The research problem is to achieve optical image edge detection without digital processing’s size, speed, and power demands or Fourier optics’ bulky components. The paper designs dielectric-nanobeam metasurfaces with engineered momentum-space transfer functions for spatial differentiation. The differentiation response agrees with the numerical calculation shown in Figure 2.
Problem
The research problem is to achieve optical image edge detection without digital processing’s size, speed, and power demands or Fourier optics’ bulky components.
Source: p. 2Method
The paper designs dielectric-nanobeam metasurfaces with engineered momentum-space transfer functions for spatial differentiation.
Source: pp. 2, 8Results
The differentiation response agrees with the numerical calculation shown in Figure 2.
Source: p. 6Takeaways & Limitations
The work shows the potential of dielectric metasurfaces for all-optical edge detection.
Source: p. 8For the 1D geometry, differentiation is performed only along the x-axis.
Source: p. 4
Abstract
from arXiv · showhide
Image processing and edge detection are at the core of several newly emerging technologies, such as augmented reality, autonomous driving and more generally object recognition. Image processing is typically performed digitally using integrated electronic circuits and algorithms, implying fundamental size and speed limitations, as well as significant power needs. On the other hand, it can also be performed in a low-power analog fashion using Fourier optics, requiring however bulky optical components. Here, we introduce dielectric metasurfaces that perform optical image edge detection in the analog domain using a subwavelength geometry that can be readily integrated with detectors. The metasurface is composed of a suitably engineered array of nanobeams designed to perform either 1st- or 2nd-order spatial differentiation. We experimentally demonstrate the 2nd-derivative operation on an input image, showing the potential of all-optical edge detection using a silicon metasurface geometry working at a numerical aperture as large as 0.35.
Methods
The methods describe fabrication of the c-Si metasurface and separate resist-patterned images with chromium lift-off.
Metasurface fabrication
- c-Si on Al2O3 substrates were cleaned and etched to the final metasurface thickness using a two-step RIE process with Cl2, HBr, and O2.The c-Si substrates were acquired from MTI Corp.Source: Methods · p. 9
- A 200 nm CSAR 62 resist layer was spin-coated and baked before Raith Voyager exposure and development; the pattern was transferred into c-Si by two-step RIE.The resist was exposed at 50 kV with a dose of 145–150 μC/cm2, then developed in Pentyl-acetate and o-Xylene.Source: Methods · p. 9
Image fabrication
- Images were patterned on cleaned 24×24 mm glass slides using a baked MMA/PMMA bilayer, Raith Voyager exposure, and MIBK:IPA development.Exposure used 50 kV and a dose of 550 μC/cm2; development used MIBK:IPA at 1:3 for 90 seconds.Source: Methods · p. 9
- A 40 nm Cr layer was thermally evaporated onto the patterned images, and residual resist was lifted off in anisole at 65˚C.Source: Methods · p. 9