Holographic Display Improvements Ready to Enhance Virtual and Augmented Reality
Advances in software and hardware can make holography usable for more applications.
Researchers have developed a new approach that improves the quality and contrast of images for holographic displays. This new technology could help improve the close-eye look used for virtual and augmented reality applications.
“Augmented reality and virtual reality systems are ready to have a transformative impact on our society by providing a seamless interface between users and the digital world,” said research team member Jonghyun Kim of technology companies NVIDIA and Stanford University. “Holographic displays can overcome some of the biggest remaining challenges for the system by improving the user experience and enabling a more compact device.”
In Optica, the journal of The Optical Society (OSA) for high-impact research, researchers described their new holographic display technology called Michelson holography. This approach combines new optical settings inspired by Michelson interferometry with the latest software development. The setting produces the interference pattern required to create a digital hologram.
The unfifififiable light of the two SLM naturally creates a fringe pattern. The camera-in-the-loop algorithm iteratively optimizes both phase patterns to create target images. Credit: Jonghyun Kim, Nvidia, Stanford University
“While we’ve recently seen remarkable advances in machine learning-driven computer-generated holography, these algorithms are essentially limited by the underlying hardware,” Kim said. “We designed together new hardware configurations and new algorithms to address some of these limitations and demonstrate sophisticated results.”
Improving holographic display quality has the potential to outperform other 3D display technologies used for virtual and augmented reality by enabling a more compact display, improving the user’s ability to focus their eyes at different distances and offering the ability to customize for users wearing corrective lenses. However, the technology has not achieved image quality like more conventional technology.
For holographic displays, image quality is limited by optical components known as phase-specific spatial light modulators (SLM). SLM creates diffraction light that makes interference patterns necessary to form visible 3D images. However, only phase SLM typically used for holography shows low diffraction efficiency which significantly lowers the observed image quality, especially image contrast.
Because it is difficult to dramatically improve the efficiency of SLM diffraction, the researchers designed a completely new optical architecture to create holographic images. Instead of using SLM only a single phase like most setups, michelson holographic approach they use SLM only two phases.
“The core idea of Michelson’s holography is to destructively disrupt the diffraction of light one SLM using unfistructed light from the other SLM,” Kim said. “This allows unfifided light to contribute to forming images rather than creating spots and other artifacts.”
Optimizing images The researchers combined these new hardware settings with the camera-in-loop optimization (CITL) procedures they modified for their optical settings. CITL optimization is a computational approach that can be used to optimize holograms directly or to train computer models based on neural networks.
CITL allows researchers to use cameras to capture a series of images shown. This means they can correct minor inequalities in optical systems without using the right measuring device.
“Once a computer model is trained, it can be used to precisely know what the image is like without physically taking it,” Kim said. “This means that all optical settings can be simulated in the cloud to perform real-time inferences from severe computational problems with parallel computing. This can be useful for calculating computer-generated holograms for complex 3D scenes, for example. ”
The researchers tested their new Michelson holographic architecture using benchtop optical settings in their lab. They used it to display some 2D and 3D holographic images, recorded with conventional cameras. The demonstration showed that the dual SLM holographic display with CITL calibration provides much better image quality than the existing computer-generated hologram approach.
For this new system to be practical, it requires translation of benchtop settings into a system small enough to be incorporated into a wearable augmented reality or virtual reality system. The researchers point out that their approach to designing hardware and software could be useful for improving other applications of computing display and computational imaging in general.