Performance parameters for detection in low-flux coherent images

Area of Science:

• Optics and Photonics
• Signal Processing
• Statistical Inference

Background:

• Coherent active imaging systems are crucial for target detection.
• Optimizing system performance in low-flux conditions presents significant challenges.
• Traditional metrics may not be optimal for unknown target and background parameters.

Purpose of the Study:

• To optimize low-flux coherent active imagery systems for enhanced target detection.
• To evaluate the efficacy of different figures of merit for detection algorithms.
• To determine the optimal pulse energy distribution for improved detection capabilities.

Main Methods:

• Analysis of optimization strategies for low-flux coherent active imagery.
• Comparison of Bhattacharyya distance and Fisher ratio as figures of merit.
• Application of generalized likelihood ratio test (GLRT) based detection algorithms.
• Simulation and analysis of target detection performance under varying pulse energy configurations.

Main Results:

• Bhattacharyya distance is demonstrated as a more efficient figure of merit than the Fisher ratio for GLRT-based detection when target and background means are unknown.
• Optimal target detection capabilities are achieved when pulse energy is divided into four shots.
• Utilizing more than ten shots for pulse energy distribution yields no significant improvement in detection results.

Conclusions:

• Bhattacharyya distance provides a robust metric for optimizing low-flux coherent active imaging systems.
• A four-shot pulse energy strategy is recommended for maximizing detection performance.
• Further increases in pulse energy beyond ten shots offer marginal benefits for this system configuration.