Optimization of light fields in ghost imaging using dictionary learning.

Area of Science:

• Optics and photonics
• Computational imaging
• Signal processing

Background:

• Ghost imaging (GI) relies on light field correlations for object reconstruction.
• Traditional GI methods suffer from poor image quality due to limited sampling.
• Reconstruction quality in GI is fundamentally limited by modulated light fields.

Purpose of the Study:

• To enhance ghost imaging (GI) quality by optimizing light fields.
• To address limitations of traditional GI reconstruction methods.
• To improve imaging performance at low sampling rates.

Main Methods:

• Developed a novel ghost imaging (GI) scheme by optimizing light fields.
• Employed matrix optimization with a learned dictionary incorporating object sparsity prior.
• Derived a closed-form solution for the sampling matrix enabling successive sampling.

Main Results:

• The proposed scheme significantly improves ghost imaging (GI) quality.
• Demonstrated superior performance compared to state-of-the-art light field optimization methods.
• Achieved better imaging results, particularly at low sampling rates.

Conclusions:

• Optimizing light fields is a viable strategy to enhance ghost imaging (GI) quality.
• The proposed matrix optimization method offers a significant advancement for GI.
• This approach provides improved imaging performance, especially in data-limited scenarios.