An adaptive and dynamic lens-flare-free single-pixel imaging method for spacecraft docking

Lens flare, the photographic artifact produced by reflecting and scattering a powerful incident light within a camera lens group, can severely degrade image quality and mislead the image-based processing task. Existing solutions typically depend on sophisticated optics design or complex image postprocessing algorithms to reduce lens flare. This paper demonstrates a direct reconstruction of the target feature by single-pixel imaging with a simple optics design and no image postprocessing. A novel sampling strategy is proposed for spacecraft docking, which utilizes Fourier single-pixel imaging methods to capture a specific feature at an ultralow sampling ratio. Simulation and experiments demonstrate that the proposed method can image the feature with a sampling ratio of 4.8% at a video frame rate while almost eliminating the impact of the ring effect. This paper provides a new perspective for achieving lens-flare-free, video-framerate and low-data-throughput imaging in space tasks, which has potential in many imaging-based applications.

Automated summary

This paper presents a method for directly reconstructing target features using single-pixel imaging, which can achieve lens-flare-free imaging without the need for image postprocessing. By utilizing a novel sampling strategy and Fourier single-pixel imaging methods, specific features can be captured at an ultra-low sampling ratio.