SAR Coherent Change Detection for Moving Target Shadow Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current synthetic aperture radar (SAR) systems face challenges in detecting moving targets and their shadows due to limitations in coherent change detection (CCD) imagery, including false alarms, inability to estimate target motion parameters, and the need for multiple imaging passes, which results in inadequate temporal resolution for tracking rapid dynamics.
Innovation Solution
A method involving the processing of first and second temporal sequences of SAR returns from separate imaging passes to form coherent and incoherent change detection images, allowing for the detection and tracking of moving targets by analyzing vehicle trails and shadow changes, independent of the target's radar cross-section and motion parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coherent change detection (CCD) is used to detect moving targets, then detection capability is improved, but false alarms increase and temporal resolution deteriorates
Solution Approach 1:
The patent segments the detection process into multiple independent imaging passes, each contributing to the construction of change detection images. By dividing the temporal sequence into discrete passes and processing them separately before integration, the system reduces false alarms while maintaining detection capability. Each pass independently captures scene changes, and the segmentation allows for better control of detection parameters across different time points.
Solution Approach 2:
The patent performs preliminary processing of SAR returns from multiple passes before final change detection image generation. Motion parameters are estimated and compensation is applied in advance to account for target movement between passes. This preliminary action prepares the data to reduce false alarms caused by motion-induced coherence loss, while the actual detection occurs in the final integrated image.
2Measurement precision
If multiple imaging passes are used to improve temporal resolution, then detection accuracy is improved, but processing time and system complexity increase
Solution Approach 1:
The patent employs a universal processing framework that handles multiple imaging passes through a single integrated change detection algorithm. The same processing pipeline processes returns from any number of passes, making the system multi-functional without requiring separate processing paths. This universality reduces system complexity despite handling multiple passes, as the core detection logic remains consistent across different numbers of inputs.
Solution Approach 2:
The patent implements dynamic estimation of target motion parameters from the temporal sequence of SAR returns. Rather than using fixed parameters, the system adapts motion estimates based on the actual target behavior observed across passes. This dynamic approach allows the system to efficiently process variable numbers of passes without requiring complex pre-programmed handling for each scenario, reducing overall system complexity.
3Measurement precision
If CCD processing is applied to detect target shadows, then shadow detection capability is improved, but inability to estimate motion parameters persists
Solution Approach 1:
The patent introduces an intermediary processing step that extracts motion parameters from the change detection images themselves. Rather than relying solely on shadow position changes, the system uses the shadow regions in CCD images as intermediate data to infer target motion characteristics. This intermediary approach recovers motion information that would otherwise be lost in shadow-only detection methods.
Solution Approach 2:
The patent implements feedback loops where estimated motion parameters from preliminary processing are used to refine shadow detection, and detected shadow positions feed back to improve motion parameter estimation. This iterative feedback process ensures that shadow detection and motion parameter estimation mutually reinforce each other, eliminating the information loss that would occur in single-pass methods.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables accurate detection and location of moving targets with reduced false alarms and improved temporal resolution, allowing for real-time observation of dynamic changes, independent of the target's radar signature and motion parameters, and provides focused images of the targets.
Implementation Method 1
SAR as an imaging technique has been developed to obtain high resolution radar imagery of surface features. It uses a technique of coherently integrating samples collected from a moving platform over a period of time
Implementation Method 2
Synthetic Aperture Radar (SAR) systems are known. SAR as an imaging technique has been developed to obtain high resolution radar imagery
Implementation Method 3
For best results these samples have to be all aligned in phase for a SAR image to be properly focused in azimuth. For example, in order to obtain a focused image of the static ground from a sideways looking SAR this is simply a quadratic phase correction across a nominally straight-line synthetic aperture
Implementation Method 4
A pair of SAR images collected for the same scene at different time instances can then be compared against each other to locate any changes in the scene that have occurred in the interval between collections
Implementation Method 5
ICD merely identifies the changes in the mean backscatter power of the scene. Typically, the average image intensity ratio of the image pair is computed to detect such changes
Implementation Method 6
CCD on the other hand, identifies changes in both the amplitude and phase content between image pairs. It relies upon the imaging processes being coherent and requires the SAR images as complex data i.e. where each pixel has a real and imaginary value. Specifically, CCD will not work with modulus of the SAR images or equivalent since the phase information is lost in the image transformation process. The phase change between a pair of images can be computed using their sample coherence
Implementation Method 7
CCD is also able to reveal phase changes resulting from the masking of the background clutter return due to obscuration from an object present in the scene, i.e. a shadow region. In such circumstances no clutter signal will be recorded from the shadow region and hence the pixel value will be merely the thermal noise value produced by the receiver. This means that the clutter intensity value is substituted for a low value determined by the thermal noise floor and the clutter phase value is replaced by a random phase. It is this latter alteration that then results in a loss of coherence that can be observed in the CCD image
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a system and method for processing imagery, such as may be derived from a coherent imaging system e.g. a synthetic aperture radar (SAR). The system processes sequences of SAR images of a region taken in at least two different passes and generates Coherent Change Detection (CCD) base images from corresponding images of each pass. A reference image is formed from one or more of the CCD base images images, and an incoherent change detection image formed by comparison between a given CCD base image and the reference image. The technique is able to detect targets from tracks left in soft ground, or from shadow areas caused by vehicles, and so does not rely on a reflection directly from the target itself. The technique may be implemented on data recorded in real time, or may be done in post-processing on a suitable computer system.