Standardized target-specific detectivity metric for computational imaging systems.

Area of Science:

• Computational imaging
• Optical engineering
• Image processing

Background:

• Multifunctional cameras performing simultaneous imaging tasks are emerging.
• Computational imaging (CI) is a key enabling technology for these cameras.
• Quantifying CI system performance is challenging due to non-traditional designs.

Purpose of the Study:

• To present a standardized framework and metric for evaluating CI system performance.
• To address the need for reliable performance quantification in computational imaging.

Main Methods:

• Introduction of a standardized detection signal-to-noise ratio, termed the detectivity metric.
• Development of a general CI system framework to accommodate diverse systems.
• Presentation of an analytical version of the detectivity metric for compressive sensing CI systems.

Main Results:

• The proposed detectivity metric offers a standardized way to assess CI system performance.
• The metric is flexible, handling various CI systems and targets with minimal assumptions.
• Standardization considerations for CI performance metrics are discussed.

Conclusions:

• The detectivity metric provides a standardized approach to quantify computational imaging system performance.
• This metric facilitates the comparison and specification of diverse CI systems.
• Standardized performance evaluation is crucial for the advancement of multifunctional cameras.