Ultra-adaptive holographic computational imaging for seeing through random scatter, fibers, and around corners

Optical imaging systems play an instrumental role in our modern life, from smartphones and automotive cameras to microscopes that are critical for clinical diagnostics.

However, the penetration depth of even the most advanced systems is still unbearably limited by the inherent random scattering of light in complex samples.

Examples span many applicative fields, including scattering in tissues and fog, limiting microscopes and laser-based systems.

For decades, the notion of correcting scattering seemed unfeasible since it requires control of billions of optical modes.

This conception changed a decade ago with the paradigm-shifting revolution of wavefront-shaping, demonstrating that scattering can be physically corrected using spatial light modulators with a number of pixels orders-of-magnitude smaller than the number of scattered modes.

Wavefront-shaping led to astonishing breakthroughs, including my own works, from focusing through visually-opaque barriers to imaging around corners.

However, beyond laboratory demonstrations, there is a fundamental gap in applying these revolutionary notions in most practical imaging applications, as wavefront-shaping is based on a physical, inherently 2D, limited-speed correction to a volumetric dynamic scattering problem, and it relies on known ‘guide-stars’ at the target.

I propose a radically different route to remove these fundamental barriers and unleash the full applicative potential of wavefront-shaping, by shifting the burden from the physical hardware to a digital, naturally-parallelizable computational process.

My approach is based on computationally emulating the optimal 3D wavefront-correction, found using only a few unique rapid holographic measurements.

My solution is enabled by our recent discovery of guide-star free wavefront-shaping, where the target themselves serve as guide-stars, and the great increase in computational power. Its impact spans across important domains, from endoscopy to non-line-of-sight imaging.