Under-sampled retina-inspired Hadamard ghost imaging enhanced by cake-cutting ordering

Area of Science:

• Optics and Imaging Technologies
• Biomimetic Systems
• Computational Imaging

Background:

• Ghost imaging (GI) utilizes correlations between a reference beam and an object beam to reconstruct images.
• Retina-inspired imaging techniques enhance efficiency by focusing on regions of interest, mimicking human visual perception.
• Hadamard speckle patterns offer rapid generation and inverse transformation capabilities for GI.

Purpose of the Study:

• To develop an efficient, under-sampled ghost imaging technique inspired by the human retina.
• To enhance image resolution and reduce computational cost using novel speckle pattern ordering and optimization.
• To integrate biomimetic principles with Hadamard transforms for advanced ghost imaging.

Main Methods:

• Implementation of under-sampled retina-inspired Hadamard ghost imaging.
• Application of cake-cutting ordering for speckle pattern arrangement based on 2D connected components.
• Incorporation of lateral inhibition enhancement for optimizing low-contrast imaging.

Main Results:

• Imaging results transition from coarse to fine with increasing sampling rates, mimicking human visual perception.
• Demonstrated high-resolution imaging of targets at low sampling rates and reduced computational load.
• Successful integration of information compression, rapid Hadamard transforms, and efficient ordering.

Conclusions:

• The proposed method effectively combines biomimetic information compression, Hadamard speckle efficiency, and cake-cutting ordering.
• Achieved high-resolution imaging with low sampling and computational costs, complementing existing retina-inspired GI.
• Shows significant promise for advancing ghost imaging technology, particularly for efficient target capture.