Sharper Images Emerge As Quantum Noise Is Optically Reduced

Li Gong and colleagues at University of Oxford, in collaboration with Quantum Innovation Centre (Q. InC), have developed a new technique to overcome fundamental limits in optical imaging resolution by enhancing the detection of high spatial frequencies obscured by noise. Their passive imaging method, Fourier Domain Division (FDD), optimises quantum measurement by pre-processing incoming light and partitioning the Fourier plane for independent detection. This approach reduces the number of photons needed to achieve a specific signal-to-noise ratio, resulting in a five-fold improvement in Fisher information for high spatial-frequency components and offering enhanced resolution in low-light conditions. Importantly, as a passive technique, unlike active super-resolution methods, FDD broadens the scope of imaging possibilities to areas where active illumination is not feasible, such as astronomy and remote sensing.

Enhanced sensitivity via Fourier plane partitioning overcomes fundamental limits to optical

Conventional microscopy’s capabilities are exceeded by a factor of five in high spatial-frequency components, achieved through Fourier Domain Division (FDD). This advancement allows the detection of details previously obscured by shot noise, a limitation stemming from the quantum nature of light and previously insurmountable with standard techniques. FDD optically pre-processes light, partitioning the Fourier plane, a representation of spatial frequencies, for independent detection and subsequent image reconstruction. Comparable signal-to-noise ratios in the Fourier domain were achieved with a five-fold reduction in required photons, enhancing sensitivity. This improvement was validated using a fluorescent resolution target, a standard test sample in microscopy.

The team employed a generalised Wiener filter, a computational technique, to combine data from the partitioned Fourier plane, maximising information retrieval. Furthermore, the passive nature of Fourier Domain Division, or FDD, extends its application beyond microscopy, offering potential use with astronomy’s extremely large telescopes where active illumination is impossible due to sample sensitivity. Experimental results demonstrated a 1.1-fold resolution improvement, though current efforts focus on optimising high spatial frequency components.

A complete assessment of the method’s performance across the entire image spectrum, and its scalability to complex three-dimensional samples, remains to be established. By cleverly manipulating light’s spatial frequencies, akin to a prism separating white light into a rainbow, Fourier Domain Division allows detailed analysis of the light’s structure. The technique partitions the Fourier plane into distinct regions, enabling independent detection of these components instead of processing them as a whole. This is important because conventional imaging systems act as a low-pass filter, suppressing high spatial frequencies and creating a blurred image, particularly when hampered by shot noise, the random fluctuations resembling static on a television screen.

Passive super-resolution imaging preserves delicate samples without active illumination

The fundamental limits of light have long hampered the quest for clearer images. Conventional optical systems struggle to resolve fine details when photons are scarce, a problem exacerbated by the inherent ‘noise’ of quantum light. Unlike competing active illumination methods, this new technique, Fourier Domain Division, avoids the need to illuminate the sample. Active approaches force extra photons onto the subject, potentially damaging delicate specimens or distorting the observation, while FDD achieves resolution gains passively, a significant advantage.

Preserving specimen integrity and data accuracy makes this particularly valuable in fields like biological microscopy, astronomy, and remote sensing where active illumination is either impossible or undesirable. A new approach to optical imaging is now available, demonstrating a pathway to resolving finer details with reduced light requirements. Fourier Domain Division cleverly partitions light’s spatial frequencies, enabling independent detection and subsequent image reconstruction, overcoming limitations imposed by conventional microscopy’s low-pass filtering and associated shot noise. Avoiding active illumination, this passive method broadens imaging possibilities to sensitive applications like astronomy and biological studies where illumination could prove damaging.

The research demonstrated a five-fold improvement in Fisher information on high spatial-frequency components using a technique called Fourier Domain Division. This method enhances resolution in low-light conditions by optimising how light is detected and processed, rather than simply increasing illumination. By partitioning light’s spatial frequencies, the technique overcomes limitations of conventional microscopy and reduces the impact of quantum shot noise. This passive approach preserves specimen integrity, making it suitable for applications where active illumination is impractical, such as astronomy and biological imaging.