SAR Object Coherence Mapping for Small Complex Geometries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for creating interferometric coherence data products for objects imaged by Synthetic Aperture Radar (SAR) face challenges such as speckle noise, inaccurate coherence estimation for small and complex-shaped objects, and unreliable data due to low geolocation accuracy and resolution, leading to distorted and unreliable coherence data products.
Innovation Solution
A method and system that determine coherence values and statistics for identifiable object geometries using geocoded interferometric SAR images, considering known geolocation accuracy and resolution, and employing adaptive windowing techniques to filter out irrelevant pixels, ensuring accurate and reliable coherence data production for small and complex-shaped objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If spatial averaging is done using a local neighborhood window to estimate coherence, then processing is simplified, but coherence estimation becomes distorted for small and complex-shaped objects because SAR data inherently contain speckle and the window includes pixels outside objects of interest
Solution Approach 1:
The patent applies local quality by using object masks to define precise spatial boundaries for each object of interest. Instead of applying a uniform local neighborhood window across the entire SAR image, the method creates object-specific regions of interest that adapt to the shape, size, and location of individual objects. This allows coherence estimation to be performed only within the actual object boundaries, excluding irrelevant pixels and speckle noise from surrounding areas, thereby maintaining measurement precision while simplifying processing through targeted analysis.
2Productivity
If existing techniques are used to produce coherence data products for objects with low absolute location accuracy and low SAR resolution, then processing is faster, but the coherence data products become inaccurate and unreliable
Solution Approach 1:
The patent applies preliminary action by performing geolocation correction and object mask creation before coherence estimation. The method pre-processes the SAR data by correcting geolocation errors, creating accurate object masks that account for low resolution and location accuracy issues, and defining precise regions of interest in advance. This preliminary preparation ensures that subsequent coherence estimation is performed on correctly positioned and bounded data, maintaining reliability without sacrificing productivity through efficient pre-processing steps.
3Device complexity
If a fixed window size is used for coherence estimation, then processing is simpler, but meaningful coherence data cannot be obtained for objects of varying sizes and complex geometries
Solution Approach 1:
The patent applies dynamics by implementing adaptive windowing based on object characteristics. Instead of using a fixed window size, the method dynamically adjusts the region of interest for each object based on its detected shape, size, and geometry. The object masks are created to match the specific boundaries of each object, allowing the coherence estimation to adapt to varying object types, sizes, and complexities. This dynamic approach maintains algorithmic simplicity through automated mask generation while achieving high versatility across different object types.
Data Source
Figure 1
Figure 2~3
Figure 4~5B
AI summary
Disclosed is a method for creating interferometric coherence data products for objects imaged by Synthetic Aperture Radar (SAR) having polarization(s). The method comprises: for each identifiable object geometry for which there are acquired geocoded interferometric SAR images with flat-earth and topographic phase removed having following pixel values: backscatter intensity (V1) in polarization(s) for master image; backscatter intensity (V2) in polarization(s) for slave image; inphase component (V3) of geocoded interferogram in polarization(s); and quadratic-phase component (V4) of geocoded interferogram polarization(s), determining which geocoded pixels are within identifiable object geometry considering known geolocation accuracy of identifiable object geometry and geocoded interferometric SAR images; and determining coherence values and statistics of coherence values for objects for polarization(s), based on pixel values V1, V2, V3, and V4 within identifiable object geometry.