Enhanced Fourier single-pixel imaging via positive-negative dithering.

Area of Science:

• Optics and Photonics
• Computational Imaging
• Digital Signal Processing

Background:

• Fourier single-pixel imaging (FSI) utilizes digital micromirror devices for efficient imaging.
• Current FSI methods employ upsampling and spatial dithering, which can reduce spatial resolution.
• There is a need for improved FSI techniques that maintain high spatial resolution.

Purpose of the Study:

• To propose and validate a novel binarization method for FSI.
• To enhance reconstructed image quality without sacrificing spatial resolution.
• To reduce quantization errors in Fourier coefficient acquisition.

Main Methods:

• A new binarization method applying spatial dithering with a serpentine path to positive and negative Fourier pattern components.
• Quantization of Fourier components into {-1, 0, +1} values.
• Mapping quantized components to binary patterns without upsampling.

Main Results:

• Significant improvement in reconstructed image quality demonstrated through simulations and experiments.
• Preservation of spatial resolution compared to traditional upsampling methods.
• Reduced quantization errors in Fourier coefficient acquisition.

Conclusions:

• The proposed binarization method enhances FSI imaging quality without compromising spatial resolution.
• This technique offers a viable alternative to upsampling in FSI.
• The method is applicable to other single-pixel imaging techniques using positive-negative grayscale patterns.