Scientists have demonstrated a new way to capture ultra-sharp optical images without lenses, precision alignment, or traditional optical constraints. The technique, developed at the University of Connecticut, replaces bulky optics with computation, opening a path to wide-field, sub-micron imaging from distances once considered impossible, according to Science Daily.
The work, led by Professor Guoan Zheng, builds on ideas borrowed from radio astronomy, where networks of separated sensors work together to simulate a much larger instrument. That same principle famously enabled the Event Horizon Telescope to capture the first image of a black hole. Translating that approach to visible light, however, has long been considered impractical due to the extreme precision normally required to keep optical signals synchronized.
The new system, called the Multiscale Aperture Synthesis Imager, or MASI, takes a different route. Instead of forcing sensors to stay perfectly aligned during measurement, each sensor operates independently. They record raw diffraction patterns describing how light waves scatter after interacting with an object. Crucially, synchronization happens later, in software, rather than during data capture.
This software-first strategy is what allows MASI to sidestep decades-old physical limits. At optical wavelengths, even nanometer-scale misalignments can destroy coherence. MASI avoids that problem by reconstructing the full wavefield at each sensor and then computationally correcting phase differences between them. Through iterative optimization, the system merges these independent measurements into a single, highly coherent image.
Another radical departure is the complete removal of lenses. Traditional imaging relies on glass to bend and focus light, introducing trade-offs between resolution, field of view, and working distance. MASI uses no lenses at all. Instead, it places coded sensors within a diffraction plane and reconstructs the image numerically. The result is a virtual synthetic aperture far larger than any individual sensor.
That virtual aperture enables sub-micron resolution across a wide field, even when the object is centimeters away from the sensor array. In conventional optics, achieving that level of detail would require placing a high-powered lens extremely close to the subject, often making the setup impractical or invasive.
Researchers say the implications extend well beyond the lab. Potential applications range from biomedical imaging and forensic analysis to semiconductor inspection and remote sensing. Because the system scales linearly as sensors are added, it could grow into large arrays without the exponential complexity that plagues traditional optical instruments.
Rather than refining lenses or chasing tighter tolerances, MASI demonstrates a different philosophy altogether. By letting computation handle what physics makes difficult, the technology suggests a future where optical imaging is defined less by glass and alignment, and more by algorithms and scalable sensor networks.
