Optimization of thermal ghost imaging: high-order correlations vs. background subtraction

Area of Science:

• Quantum optics
• Computational imaging
• Image processing

Background:

• Thermal ghost imaging is a technique that uses correlated thermal light beams to form an image.
• High-order thermal ghost imaging extends this concept to higher-order correlations, potentially offering improved performance.
• Conventional ghost imaging typically uses lowest-order correlations.

Purpose of the Study:

• To compare the performance of high-order thermal ghost imaging with conventional thermal ghost imaging.
• To evaluate different data processing methods, specifically background normalization for high-order imaging and background subtraction for conventional imaging.
• To determine the optimal data processing strategy for thermal ghost imaging.

Main Methods:

• Analytical comparison of contrast-to-noise ratio (CNR) for different thermal ghost imaging orders and data processing techniques.
• Focus on high-order thermal ghost imaging with background normalization.
• Focus on conventional (lowest-order) thermal ghost imaging with background subtraction.

Main Results:

• The CNR of normalized high-order ghost images scales inversely with the square root of the number of transmitting object pixels.
• This scaling law is independent of the order of correlation used in high-order thermal ghost imaging.
• No data processing method evaluated outperformed conventional ghost imaging with background subtraction.

Conclusions:

• Conventional ghost imaging with background subtraction remains the most effective method among those studied.
• The analytical findings provide a theoretical basis for understanding the performance limitations and scaling laws in thermal ghost imaging.
• The results help explain experimental observations in recent ghost imaging studies.