Differenced Change Product for SAR Clutter Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Coherent Change Detection (CCD) images generated from Synthetic Aperture Radar (SAR) systems often contain clutter, making it difficult for analysts to identify human-induced changes in a scene, as low coherence regions can be attributed to both human activity and environmental factors like vegetation or shadows.
Innovation Solution
The generation of a Differenced Change Product (DCP) image, which is derived from a stack of CCD images, involves scaling pixel values using a monotonic mapping function to diminish differences in low coherence regions, allowing for clearer identification of human-induced changes by highlighting areas with significant coherence differences between SAR image pairs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If CCD images are generated from SAR imagery to detect scene changes, then change detection capability is improved, but image clutter increases making analysis difficult
Solution Approach 1:
The patent segments the change detection process into multiple temporal passes, comparing SAR imagery across three or more time points. By dividing the analysis into sequential comparisons (first pass vs. second pass, second pass vs. third pass, etc.), the system isolates human-induced changes from environmental clutter, allowing analysts to track changes over time while filtering out persistent noise sources like vegetation and shadows.
Solution Approach 2:
The patent performs preliminary processing by generating multiple CCD images from sequential SAR passes before final analysis. These pre-processed coherence magnitude images serve as prepared inputs that highlight potential changes, which are then further analyzed through temporal comparison to distinguish genuine human activity from environmental factors, reducing the need for extensive post-processing.
2Measurement precision
If multiple SAR passes are processed to improve change detection, then detection accuracy is improved, but processing time increases
Solution Approach 1:
The patent segments the multi-pass processing into efficient sequential steps: generating coherence magnitude images for each pass pair, then comparing these segmented results to identify consistent changes. This segmentation allows parallel processing of individual pass comparisons while maintaining temporal context, reducing overall processing time compared to analyzing all passes simultaneously.
Solution Approach 2:
The patent applies change detection to specific passes or time windows rather than processing all available SAR imagery. Analysts can selectively process only the passes most relevant to the investigation, performing partial processing that balances detection accuracy with time constraints, avoiding the need to analyze every available pass when fewer passes suffice.
Data Source
AI summary
Described herein are various technologies relating to constructing a differenced change product (DCP) image. A plurality of synthetic aperture radar (SAR) images of a scene are generated based upon radar signals directed towards and reflected off of the scene, and a plurality of coherence change detection (CCD) images of the scene are generated based upon the SAR images. The CCD images are registered with one another, and their pixel values re-scaled according to a monotonic mapping function. The DCP image is generated based upon a computed pixel-wise difference between a pair of the re-scaled CCD images. The DCP image identifies locations in the scene where human activity-induced change is likely to have occurred between a pair of SAR passes of the scene.