Full-chain model-driven cross-level multi-parameter collaborative optimization for electro-optical imaging systems

Area of Science:

• Optics and Photonics
• Image Processing
• Systems Engineering

Background:

• Current infrared imaging system optimization relies on linear models, neglecting non-ideal factors and limiting performance prediction accuracy.
• Detection and recognition of infrared targets in complex environments require advanced imaging capabilities.
• Existing frameworks lack a unified approach to incorporate nonlinear coupling and non-ideal factors.

Purpose of the Study:

• To establish a full-chain physical modeling framework for infrared electro-optical imaging systems.
• To enable unified characterization of nonlinear coupling among spectral response, spatial transfer, sampling, and noise.
• To develop a collaborative optimization model for high-performance infrared imaging system design.

Main Methods:

• Developed a full-chain physical modeling framework from radiative input to grayscale image output.
• Adopted simulated image contrast as the objective function for optimization.
• Employed a hybrid optimization strategy combining genetic algorithms and local search.

Main Results:

• Coupled structural parameters and non-ideal imaging factors within a unified performance evaluation space.
• Achieved stable convergence and significant contrast improvement in simulations.
• Demonstrated strong cross-scene adaptability for land and sea clutter environments.

Conclusions:

• The proposed framework provides a systematic approach for digital collaborative design of infrared imaging systems.
• The unified model enhances performance prediction reliability by incorporating non-ideal imaging factors.
• Optimized systems show improved detection and recognition capabilities in complex environments.