Super-Resolution Image Display Using Diffractive Decoders (Case No. 2022-302)

Summary

UCLA researchers in the Department of Electrical and Computer Engineering have developed a super-resolution (SR) image display framework that increases the effective number of useful pixels by 16-fold. This framework combines deep learning-based digital encoding with all-optical decoding to achieve ultra-fast, energy-efficient image super-resolution for advanced display and communication applications.

Background

Virtual reality (VR) and augmented reality (AR) are poised to transform multiple industries, offering immersive experiences that blend physical reality with digital content. However, traditional AR/VR systems face challenges in delivering high image quality while maintaining power efficiency and compact form factors. Existing holographic and optical imaging approaches often suffer from high power consumption, memory burden, computational overhead, and complexity of system design. To enable next-generation AR/VR systems, there is a need for a novel display framework that delivers high-resolution imaging while remaining energy- and computation-efficient.

Innovation

Professor Aydogan Ozcan and his team have invented a diffractive super-resolution image display framework that integrates a convolutional neural network-trained digital encoder with an all-optical decoder (fabricated using 3D printing). The encoder and decoder are jointly trained to enhance image fidelity, achieving a 4× super-resolution factor in both lateral dimensions—equivalent to a 16-fold increase in the effective pixel count. Because image projection occurs all-optically as light passes through the thin diffractive decoder layers, the system provides ultra-fast, energy-efficient image generation with minimal computational resources. The framework significantly reduces data transmission and storage demands and can be scaled across different wavelength regimes. This technology is applicable to display systems and free-space optical communications, even under conditions with occlusions.

Advantages

  • Produces super-resolved images with up to a 16-fold increase in effective pixels

  • Reduces data transmission and storage requirements

  • Fast, all-optical processing at the speed of light

  • Minimizes computational resources needed

  • Compact, energy-efficient, and cost-effective design

  • Simple fabrication through 3D printing with scalability to different wavelengths

Potential Applications

  • Augmented reality (AR) and virtual reality (VR) systems

  • Holographic displays

  • Machine vision and optical computing systems

  • Next-generation 3D display platforms

  • High-data-rate free-space optical communications

Development to Date

  • First successful demonstration (proof of concept): May 2021.

Patent / Application

WO2023244949 A1 — Super-resolution image display using diffractive decoders
Super-resolution image display using diffractive decoders (WO2023244949A1)

Publications by the Inventors (Related Work)

  • Işıl, Çağatay; Mengu, Deniz; Rivenson, Yair; Ozcan, Aydogan. Super-resolution image display using diffractive decoders. Science Advances 8, no. 48 (2022): eadd3433. DOI: 10.1126/sciadv.add3433

Patent Information: