Small Mover Detection in SAR Imagery Using Excess Coherency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Detecting small moving targets in radar images is challenging due to difficulties in distinguishing them from other sources of radar energy, such as azimuth sidelobes and newly placed stationary reflectors, which complicate target detection in synthetic aperture radar (SAR) images.
Innovation Solution
The method involves generating an excess coherency factor (ECF) image by comparing two SAR images captured at different passes, applying filters to extract and enhance pixels that meet specific constraints, and using image processing techniques like row differencing and bilateral gradient enhancement to identify and isolate small moving targets from spurious signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional SAR image processing is used to detect small moving targets, then the detection process becomes complicated by spurious signals, but the complexity of the detection system increases
Solution Approach 1:
The patent segments the detection process into multiple specialized filter stages: an amplitude filter to remove stationary reflectors, a coherence filter to eliminate azimuth sidelobes, and a motion filter to isolate small moving targets. Each filter addresses a specific type of spurious signal, breaking down the complex detection problem into manageable components that collectively improve target detection accuracy without requiring a single overly complex system
Solution Approach 2:
The patent introduces an excess coherency factor (ECF) image as an intermediary representation that captures changes between sequential SAR images. This ECF image serves as a mediator that highlights moving targets while suppressing stationary clutter and artifacts, allowing the detection system to work with a simplified intermediate product rather than directly processing the complex original SAR images
2Measurement precision
If multiple filtering operations are applied to extract small moving targets, then target identification accuracy improves, but processing time increases
Solution Approach 1:
The patent applies preliminary filtering operations in a specific sequence, starting with the amplitude filter to remove obvious stationary reflectors, followed by the coherence filter to eliminate azimuth sidelobes, and finally the motion filter to isolate small moving targets. This predetermined filtering sequence eliminates unnecessary processing steps and allows each filter to work most efficiently, reducing total processing time while maintaining high target identification accuracy
Data Source
AI summary
The various technologies presented herein relate to detecting small moving entities or targets in radar imagery. Two SAR images can be captured for a common scene, wherein the scene is imaged twice from the same flight path. The first image is captured at a first instance and the second image is captured at a second instance, and differences between the two images are determined using a complex SAR change measure, excess coherency factor or DeltaC, based in part upon quantification of incoherent (or magnitude) change between the two images. A plurality of operations are performed to enable extraction of coherent change measures relating to the small moving entities from measures relating to large objects, stationary reflective structures, radar focusing artifacts, etc.