3D-printed decoder, AI-enabled image compress - EurekAlert

FINDINGS

A UCLA squad has developed a exertion for projecting high-resolution computer-generated images utilizing one-sixteenth the fig of pixels contained successful their root images. The strategy compresses images based connected an artificial quality algorithm, and past decodes them utilizing an optical decoder — a thin, translucent expanse of integrative produced utilizing a 3D printer — that is designed to interact with airy successful a circumstantial mode arsenic portion of the aforesaid algorithm. The decoder consumes nary power, which could effect successful higher-resolution displays that usage little powerfulness and necessitate little information than existent show technologies.

BACKGROUND

Projecting high-resolution 3D holograms requires truthful galore pixels that the task is beyond the scope of existent user technology. The quality to compress representation information and instantly decode compressed images utilizing a thin, transparent worldly that does not devour power, arsenic demonstrated successful the study, could assistance flooded that obstruction and effect successful wearable exertion that produces higher prime images portion utilizing little powerfulness and retention than today’s user technology.

METHOD

The strategy uses an algorithm that encodes a high-resolution representation to a lower-resolution one. The effect is simply a pixelated pattern, akin to a QR code, that is unreadable to the quality eye. That compressed representation is past translated backmost to its archetypal solution by a decoder designed to crook and unscramble the incoming light.

Testing the strategy connected images successful black, achromatic and shades of gray, the researchers demonstrated that the exertion could efficaciously task high-resolution images utilizing encoded images with lone astir 6% of the pixels successful the original. The squad besides tested a akin strategy that successfully encoded and decoded colour images.

IMPACT

The exertion could yet beryllium utilized for applications similar projecting high-resolution holographic images for virtual world oregon augmented world goggles. By encoding images utilizing a fraction of the information contained successful the archetypal and decoding them without utilizing electricity, the strategy could pb to holographic displays that are smaller, little costly and person faster refresh rates.

The exertion could look successful user electronics arsenic soon arsenic 5 years from now, according to the paper’s corresponding author, Aydogan Ozcan, Chancellor’s Professor of Electrical Engineering and Bioengineering, Volgenau Professor of Engineering Innovation astatine the UCLA Samueli School of Engineering and an subordinate manager of the California NanoSystems Institute astatine UCLA.

Other imaginable applications see representation encryption and aesculapian imaging.

AUTHORS

The co-first authors of the survey are UCLA doctoral students Çağatay Işıl and Deniz Mengu. Mona Jarrahi, UCLA’s Northrop Grumman Professor of Electrical Engineering, is simply a co-senior author. Additional authors are Yifan Zhao, Anika Tabassum, Jingxi Li and Yi Luo, each of UCLA.

JOURNAL

The survey is published in Science Advances.

FUNDING

The probe was funded by the Department of Energy.