Enhancing infrared imaging robustness against laser-induced damage: a wavefront coding and Mamba-UNET approach.

Area of Science:

• Optics and Photonics
• Infrared Imaging Systems
• Laser Damage Mitigation

Background:

• High-sensitivity infrared imaging systems are vulnerable to laser radiation, causing pixel saturation and detector damage.
• Balancing laser protection with imaging quality is a critical challenge for these systems.
• Existing protection methods may compromise image fidelity.

Purpose of the Study:

• To enhance laser protection for infrared imaging systems using wavefront coding.
• To optimize a logarithmic phase mask for superior laser diffusion and image quality preservation.
• To develop an advanced image restoration algorithm for wavefront-coded infrared images.

Main Methods:

• Wavefront coding technology was employed to diffuse incident laser energy.
• A genetic algorithm was used to design and optimize a logarithmic phase mask.
• A U-Net model integrated with a Mamba module was trained on a synthetic WFC dataset for image restoration.

Main Results:

• The wavefront coding system demonstrated approximately 100 times improvement in laser protection performance.
• The proposed decoding algorithm effectively restored image details with superior imaging quality compared to existing methods.
• The enhanced U-Net Mamba architecture improved overall image restoration performance.

Conclusions:

• Wavefront coding offers a viable solution for laser protection in long-wave infrared optical imaging systems.
• The optimized logarithmic phase mask effectively reduces laser energy density on detectors without degrading image quality.
• The developed deep learning-based decoding algorithm significantly enhances image restoration capabilities for protected infrared systems.