Experimentally measuring a detectability index of a computational imaging system.

Area of Science:

• Computational imaging
• System-level performance characterization
• Non-shift invariant linear systems

Background:

• Computational imaging (CI) systems enable multifunctional cameras but are complex to evaluate.
• Characterizing the performance of complex CI systems, especially non-shift invariant linear ones, is challenging.
• Existing methods may not adequately assess system-level detection tasks.

Purpose of the Study:

• To propose and test a novel measurement technique for evaluating complex non-shift invariant linear CI systems.
• To characterize CI system performance using detectability indexes, specifically the average Hotelling's statistic (t^2).
• To compare the detection performance of different cameras using the developed measurement technique.

Main Methods:

• Developed a novel measurement technique based on a general CI system framework.
• Utilized detectability indexes, including the average Hotelling's statistic (t^2), to predict signal-to-noise ratio.
• Validated the technique by comparing experimental results with the Night Vision Integrated Performance Model (NV-IPM) and Monte Carlo simulations.
• Demonstrated the technique on various target sizes, colors, and brightnesses against different backgrounds.

Main Results:

• The proposed measurement technique successfully evaluated complex CI systems at the system level.
• Detectability indexes provided valuable insights into the final system performance.
• Experimental results aligned with theoretical predictions from NV-IPM and Monte Carlo simulations.
• The technique effectively differentiated the detection performance between two tested cameras.

Conclusions:

• The novel measurement technique is effective for characterizing the performance of complex non-shift invariant linear CI systems.
• Detectability indexes offer a robust method for predicting and comparing CI system detection capabilities.
• This approach facilitates a deeper understanding and comparison of imaging system performance.