‘Few-pixel’ wavefront imaging mitigates drawbacks of single-pixel and array techniques

In the analysis of light waves propagating and reflecting on optical systems, conventional wavefront imaging using pixelated array detectors can underperform in certain wavelength bands and low-light environments. Wavefront single-pixel imaging (WSPI) has shown promise for capturing the wavefront of the target field with single-photon sensitivity and a broad spectral range. As imaging resolution increases, however, WSPI systems require computationally intensive modulation patterns to reconstruct high-fidelity images.

Researchers have proposed a parallel WSPI approach that boosts imaging speed by obtaining different parts of the target wavefront simultaneously. By obtaining multiplex wavefront information in parallel, the technique by Yin et al. establishes parallel common-path interferometric detections to obtain images quickly without sacrificing image quality.

“This work is the first to propose a structured ‘few-pixel’ detection framework specifically for wavefront imaging,” said author Dong Wang. “Parallel wavefront single-pixel imaging demonstrates that high-quality wavefront reconstruction is achievable using only a few strategically placed single-pixel detectors working in parallel.”

Raw parallel wavefront detection leads to crosstalk between imaging regions as light waves diffract and propagate. The team’s approach mitigates crosstalk by allocating grating arrays to sub-regions, each with a unique orientation and period, whose frequency components are spatially separated.

In simulations and experiments, four parallel interferometric detections enabled a fourfold increase in imaging speed. To confirm its applicability, the researchers used the technique to reconstruct the wavefront of a dragonfly wing.

In addition to four single-pixel detectors, the proposed architecture can be extended to 3x3 or larger grids of detectors, with imaging-speed acceleration proportional to the parallelization scale.

The group next looks to further explore more applications of “few-pixel” paradigms that balance broad spectral range, single-photon-level detection, and near-real-time acquisition capabilities.

Source: “Parallel wavefront single-pixel imaging,” by Zixian Yin, Yuxin Lu, Bochao Hu, Xinlong Liu, Aiping Zhai, and Dong Wang, Applied Physics Letters (2025). The article can be accessed at https://doi.org/10.1063/5.0274196 .