Super-Resolution Procedure for Target Responses in KOMPSAT-5 Images

Area of Science:

• Remote Sensing
• Synthetic Aperture Radar (SAR) Imaging
• Image Processing

Background:

• Satellite SAR remote sensing is crucial for target analysis, but performance is limited by the spatial resolution of target responses.
• Improving spatial resolution is essential for enhanced target analysis, yet it's constrained by system design and challenges like target motion and frequency spectrum issues.

Purpose of the Study:

• To describe a complete super-resolution procedure for target responses in KOMPSAT-5 satellite SAR images.
• To address limitations in super-resolution application to SAR images, specifically target motion-induced blurring and zero-regions in the frequency spectrum.

Main Methods:

• Extraction of target images containing target responses from large-scale KOMPSAT-5 images.
• Preprocessing of target images to mitigate target motion and optimize for super-resolution techniques.
• Application of super-resolution techniques to enhance spatial resolution and quality of target images.

Main Results:

• Demonstrated noticeable improvements in spatial resolution for various target types (static and moving) with reliable computation times.
• Enhanced performance metrics including Peak-to-Sidelobe Ratio (PSLR), Integrated Sidelobe Ratio (ISLR), and Impulse Response (IC), leading to clearer scattering information.
• Validated the practical design and performance evaluation of the super-resolution processing on real satellite SAR images.

Conclusions:

• The proposed super-resolution scheme offers a practical solution for enhancing target analysis in satellite SAR imagery.
• The method significantly improves spatial resolution and clarity of target responses, aiding in more precise target recognition.
• This work represents a first attempt at describing a whole super-resolution procedure for target responses in satellite SAR.