SAR Autofocus for Ground Penetration Radar Phase Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wide-beam/wide-band ground penetration radar (GPR) systems face challenges in achieving high azimuth resolution due to spatially varying phase errors, curved target range migration traces, and lower signal-to-noise ratios, making existing autofocus solutions ineffective.
Innovation Solution
A method that estimates target phase error directly from complex ground images using a 1D Fourier series model and contrast optimization, applying phase error correction to each image block through 2D to 1D mapping, and forming an image mosaic for improved focus.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a narrow beam SAR system is used, then one autofocus solution can be applied to the entire image, but the system cannot achieve high azimuth resolution for wide-area mapping
Solution Approach 1:
The patent divides the wide-beam SAR image into multiple sub-images or image blocks, each covering a smaller angular sector. An autofocus solution is then independently applied to each sub-image where the phase error can be considered approximately constant. This segmentation allows the system to maintain high azimuth resolution in each sub-image while collectively covering a wide area, resolving the contradiction between resolution and coverage.
Solution Approach 2:
The patent implements local autofocus correction by applying different phase error compensation parameters to different regions of the image. Instead of using a single global autofocus solution, the system calculates and applies localized phase corrections tailored to each image block's specific angular range, thereby achieving high resolution locally across the entire wide-area image.
2Measurement precision
If high performance navigation system and accurate terrain elevation knowledge are used, then high azimuth resolution can be achieved, but the system cost increases significantly
Solution Approach 1:
The patent implements autofocus functionality that enables the SAR system to self-correct its own phase errors using only the received signal data and basic navigation information. The autofocus algorithm automatically estimates and compensates for phase errors without requiring external high-precision navigation systems or detailed terrain elevation data, allowing the system to achieve high azimuth resolution while maintaining simple navigation equipment.
3Area of stationary object
If wide beam SAR is used for GPR, then the phase error becomes highly spatially varying, but this makes single autofocus solution inapplicable to the entire image
Solution Approach 1:
The patent segments the wide-beam GPR image into multiple narrow-beam sub-images, each covering a limited angular sector where phase error variation is minimal. An individual autofocus solution is computed and applied to each sub-image, ensuring phase error consistency within each segment while maintaining overall wide-area coverage through concatenation of the processed sub-images.
Solution Approach 2:
The patent applies localized phase error compensation by calculating separate autofocus parameters for different spatial regions of the GPR image. Each image block receives tailored phase correction based on its local characteristics, ensuring high measurement precision locally while collectively covering the entire ground area of interest.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A method of synthetic aperture radar autofocus for ground penetration radar (100). The method includes transmitting a signal via an antenna (102); receiving a reflected signal comprising a plurality of image blocks via the antenna (102); reading each image block from the reflected signal (602) via a processor (108); locating prominent targets in each image block (604) via the processor (108); estimating ground penetration phase error (606) via the processor (108) in each image block via phase error inputs (608) including pulling range and quantization level by generating a 1 D phase error (706) and converting the 1 D phase error into a 2 D phase error of an image spectra (708); refocusing each image block according to estimated ground penetration phase error for that image block (610) via the processor (108); and forming an image mosaic comprising each refocused image block (614) via the processor (108).