A analysis crew at Metropolis College of Hong Kong (CityUHK) has developed a brand new optical nook detection imaging technique, which makes use of azimuthal Hilbert rework metasurfaces and is designed to work as a common framework. The examine marks an necessary advance in high-speed, low-power optical data processing and has demonstrated potential for movement monitoring functions.
The examine was led by ProfessorTsai Din-ping, Chair Professor within the Division of Electrical Engineering at CityUHK. The venture was carried out in collaboration with Professor Tao Li, from the Faculty of Engineering and Utilized Sciences at Nanjing College. Titled “Optical corner detection with azimuthal Hilbert transform metasurfaces”, the examine was printed within the prestigious journal Science Advances and chosen as a highlighted work on the journal’s homepage.
The inspiration for this analysis stems from organic imaginative and prescient. Many animals are particularly good at noticing geometric options, particularly corners. For instance, bees are drawn to angular shapes and use their compound eyes to determine sharp constructions on flowers, and woodpeckers usually peck on the intersections of bark cracks or the sides of holes, the place bugs usually tend to be hidden.
These examples mirror a typical technique—extracting key data from advanced scenes by specializing in native options, reminiscent of corners, that stand out clearly. In right this moment’s world, the place data overload is frequent, this technique will help scale back visible complexity and enhance each sensing and computational effectivity.
Nook detection is extensively utilized in pc imaginative and prescient, together with picture registration, movement monitoring, and 3D reconstruction. Right this moment, most nook detection depends on digital sign processing. Nonetheless, electronics-based digital processing faces limits in response velocity and vitality consumption. Photonic or optical processing might provide sooner and extra energy-efficient efficiency, however current optical processing approaches have targeted primarily on 2D characteristic extraction. Sensible optical nook detection strategies have been restricted.
On this examine, the analysis crew constructed a 4f imaging system and launched a metasurface on the Fourier aircraft to modulate the optical subject within the angular route. They used silicon nitride because the metasurface materials and optimised the structural parameters based mostly on geometric-phase modulation, enabling environment friendly broadband operation within the seen spectrum.
They first examined metasurfaces with totally different designs, demonstrating that corners in several polygon shapes had been clearly highlighted within the photos. They confirmed secure efficiency throughout a variety of seen wavelengths, supporting the system’s broadband capabilities.
The researchers then used imaging experiments on polygon arrays to point out that the scheme may cowl the complete detection aircraft, demonstrating a large-field-of-view nook detection functionality.
They additional demonstrated {that a} single metasurface can detect objects with totally different amplitude, part, and angular traits. They analysed the imaging outcomes, deriving parity guidelines that specify how shiny and darkish factors kind within the ensuing photos.
Leveraging nook detection capabilities to extract key data, the analysis crew utilized the imaging technique to trace objects present process random rotation and translation. The experiments verified its effectiveness in supporting movement monitoring and demonstrated environment friendly data-compression functionality.
“By way of this examine, we spotlight the distinctive benefits of optical metasurfaces for real-time and environment friendly data processing,” mentioned Professor Tsai. “By offering a brand new technical pathway for all-optical nook detection, our purpose is to cut back vitality consumption and enhance processing velocity in future computational imaginative and prescient programs and next-generation clever sensing applied sciences.”
Dr Chen Chen, a postdoctoral fellow within the Division of Electrical Engineering at CityUHK, served as one of many co-first authors.
This collaborative analysis was supported by the Nationwide Key Analysis and Growth Program of China, the Nationwide Pure Science Basis of China, the Analysis Grants Council of Hong Kong, and the Areas of Excellence Scheme.
