New Sensor Rewrites Rules of Optical Imaging
Inspired by a technique that allowed astronomers to image a black hole, scientists at the University of Connecticut have developed a lensless image sensor that achieves submicron 3D resolution, promising to transform fields from forensics to remote sensing.
Operating principle and implementation of MASI. Image credit: Wang and others., doi: 10.1038/s41467-025-65661-8.
“At the heart of this breakthrough is a long-standing technical problem,” said Professor Guoan Zheng of the University of Connecticut, lead author of the study.
“Synthetic aperture imaging works by coherently combining measurements from multiple separate sensors to simulate a much larger imaging aperture.”
In radio astronomy, this is possible because the wavelength of radio waves is much longer, making precise synchronization between sensors possible.
But at visible light wavelengths, where the scale of interest is several times smaller, traditional timing requirements become almost impossible to physically satisfy.
The Multi-Scale Aperture Synthesis Imager (MASI) reverses this challenge.
Rather than forcing multiple optical sensors to work in perfect physical synchronization, MASI allows each sensor to measure light independently, then uses computer algorithms to then synchronize the data.
“This amounts to asking multiple photographers to capture the same scene, not as ordinary photos, but as raw measurements of light wave properties, and then letting software assemble these independent captures into a single ultra-high resolution image,” Professor Zheng said.
This computational phase synchronization scheme eliminates the need for rigid interferometric setups that have until now prevented the practical deployment of optical synthetic aperture systems.
MASI departs from conventional optical imaging in two transformative ways.
Rather than relying on lenses to focus light onto a sensor, MASI deploys an array of coded sensors positioned in different parts of a diffraction plane. Each captures raw diffraction patterns – essentially the way light waves propagate after interacting with an object.
These diffraction measurements contain both amplitude and phase information, which are recovered using computer algorithms.
Once the complex wavefield from each sensor is recovered, the system digitally completes and digitally propagates the wavefields toward the object plane.
A computational phase synchronization method then iteratively adjusts the relative phase offsets of each sensor’s data to maximize overall consistency and energy in the unified reconstruction.
This step constitutes the key innovation: by optimizing the combined wavefields in software rather than physically aligning the sensors, MASI overcomes the diffraction limit and other constraints imposed by traditional optics.
A virtual synthetic aperture larger than any single sensor, enabling sub-micron resolution and wide field coverage without lenses.
Conventional lenses, whether used in microscopes, cameras, or telescopes, force designers to make compromises.
To resolve smaller details, lenses must be closer to the object, often within a few millimeters, which limits the working distance and makes some imaging tasks impractical or invasive.
The MASI approach does without lenses entirely, capturing diffraction patterns from centimeters away and reconstructing the images with resolution down to submicron levels.
It’s like being able to examine the fine ridges of a human hair from a desktop computer instead of bringing it inches from your eye.
“The potential applications of MASI cover multiple fields, from forensics and medical diagnosis to industrial inspection and remote sensing,” said Professor Zheng.
“But what’s most exciting is the scalability: unlike traditional optics which become exponentially more complex as they grow, our system scales linearly, potentially enabling large arrays for applications we haven’t even imagined yet.”
The team’s paper was published in the journal Natural communications.
_____
R.Wang and others. 2025. Multi-scale aperture synthesis imager. Nat Common 16, 10582; doi: 10.1038/s41467-025-65661-8
