Compressive SAR imaging with joint sparsity and local similarity exploitation

Area of Science:

• Radar Imaging
• Signal Processing
• Image Reconstruction

Background:

• Compressive sensing (CS) enables high-resolution synthetic aperture radar (SAR) imaging.
• Existing CS-SAR methods struggle with complex scenes due to fixed sparse representations.
• Adaptive characterization of varied image content is crucial for advanced SAR imaging.

Purpose of the Study:

• To propose a novel compressive sensing-based SAR imaging approach using adaptive sparse representation.
• To enhance the adaptivity and capability of SAR imaging for complex scenes.
• To improve high-resolution image formation in SAR systems.

Main Methods:

• Introduced an autoregressive model to exploit image structural sparsity.
• Integrated pixel similarity into the autoregressive model for enhanced capability.
• Developed a weighted autoregressive model for adaptive sparse representation.
• Implemented a joint optimization scheme with iterative SAR imaging and model updating.

Main Results:

• Experimental results validated the proposed adaptive sparse representation approach.
• The method demonstrated effectiveness in high-resolution SAR image formation for complex scenes.
• The approach showed generality and validity across different imaging scenarios.

Conclusions:

• The proposed weighted autoregressive model significantly improves SAR imaging adaptivity.
• Joint optimization effectively addresses the challenge of data-driven model determination.
• This adaptive CS-SAR method offers a promising solution for complex scene imaging.