Related Experiment Video
Updated: Jul 8, 2025
08:48
Demonstration of Spin-Multiplexed and Direction-Multiplexed All-Dielectric Visible Metaholograms
Published on: September 25, 2020
5.8K
All-optical geometric image transformations enabled by ultrathin metasurfaces.
Xingwang Zhang1, Xiaojie Zhang1, Yao Duan1
1Department of Electrical Engineering, The Pennsylvania State University, University Park, PA, 16802, USA.
Nature Communications
|December 15, 2023
Summary
We developed ultra-compact metasurfaces for all-optical geometric image transformations. This technology enables efficient scale- and rotation-invariant image preprocessing for artificial visual systems.
Area of Science:
- Optics and Photonics
- Artificial Intelligence
- Image Processing
Background:
- Optical analog image processing offers parallel processing and low power consumption, crucial for artificial visual systems.
- Geometric image transformations are essential for correcting distortions, creating effects, and image morphing.
Purpose of the Study:
- To present ultra-compact metasurfaces capable of all-optical geometric image transformations.
- To demonstrate the application of these metasurfaces for binary and grayscale image manipulations.
Main Methods:
- Fabrication and characterization of ultra-compact metasurfaces.
- All-optical implementation of geometric image transformations, including Cartesian to log-polar coordinate conversion.
- Demonstration of binary and grayscale image transformations.
Main Results:
- Metasurfaces successfully performed all-optical geometric image transformations.
- Binary images were converted to log-polar coordinates, enabling scale- and rotation-invariant preprocessing.
- Grayscale images were transformed from Gaussian to flat-top intensity profiles.
Conclusions:
- The developed metasurfaces offer a powerful platform for advanced optical image processing.
- This technique has potential applications in target tracking, laser manufacturing, and biomedical imaging.
