Recursive Clutter Suppression in Video Imagery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current imaging and tracking systems face challenges in effectively suppressing clutter, especially harsh clutter, which limits sensor performance and increases noise levels, particularly in video imagery from electro-optical sensors with focal plane arrays, and complicates target detection on moving platforms.
Innovation Solution
The implementation of a recursive motion compensated integration technique that decomposes video imagery into independently moving clutter patterns, generating and subtracting successive corrections to recursively suppress these patterns, improving target detection by reducing noise and enhancing image clarity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If calibration or nonuniformity correction is used to suppress fixed pattern noise, then fixed pattern noise is reduced, but a large residual fixed pattern remains which limits sensor performance and increases sensor noise levels, especially when the raw imagery contains harsh clutter
Solution Approach 1:
The patent segments the clutter suppression process into multiple independent components: motion compensation for geometric distortion, recursive iterative processing for residual clutter, and separate handling of fixed pattern noise. This segmentation allows each component to address specific aspects of clutter without interfering with others, thereby improving overall suppression effectiveness while maintaining sensor performance.
Solution Approach 2:
The patent implements a recursive iterative processing approach where the clutter suppression algorithm is applied repeatedly until convergence is achieved. This continuous application of the suppression algorithm ensures that residual clutter patterns are progressively removed, achieving superior noise reduction compared to single-pass methods while maintaining computational efficiency.
2Object-affected harmful factors
If traditional clutter suppression techniques are applied to video imagery with harsh clutter, then some clutter is removed, but the techniques work quite well only for uncluttered or mildly-cluttered imagery and fail to effectively suppress harsh clutter
Solution Approach 1:
The patent employs a dynamic adaptive algorithm that automatically adjusts processing parameters based on the detected clutter characteristics in each frame. The motion compensation parameters and iterative suppression thresholds are dynamically optimized according to the specific clutter patterns present, enabling effective suppression across diverse conditions from uncluttered to harsh clutter environments without manual intervention.
Solution Approach 2:
The patent incorporates feedback mechanisms where the algorithm continuously monitors the effectiveness of clutter suppression and adjusts its parameters accordingly. The recursive iterative process uses feedback from previous iterations to refine subsequent suppression operations, ensuring optimal performance across different clutter conditions and enabling the system to adapt to varying environmental conditions.
3Measurement precision
If calibration is performed prior to use to suppress fixed pattern noise, then noise is reduced, but there may be instances when calibration cannot be performed prior to use and the system must be used in a moment's notice
Solution Approach 1:
The patent performs preliminary estimation of fixed pattern noise characteristics during the initial frames of operation, even before formal calibration is completed. This preliminary action enables the system to apply noise suppression techniques immediately upon receiving the first image data, eliminating the need to wait for complete calibration while still achieving effective noise reduction.
Solution Approach 2:
The patent implements self-calibrating capabilities where the system automatically estimates and corrects fixed pattern noise without requiring external calibration equipment or manual intervention. The algorithm uses the image data itself to characterize and compensate for noise patterns, enabling the system to serve its own calibration needs and operate independently in real-time scenarios.
4Difficulty of detecting and measuring
If tracking objects using an optical sensor with a telescope and focal plane array on a moving platform is performed, then target tracking is achieved, but additional problems arise such as the need to compensate for the movement of the moving platform
Solution Approach 1:
The patent merges the motion compensation functionality directly into the clutter suppression algorithm, combining what were previously separate processing stages into a unified operation. This integration eliminates the need for separate motion compensation modules while achieving both goals simultaneously, thereby reducing overall system complexity despite the added capability to handle moving platforms.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of suppressing clutter in video imagery includes receiving (402) video imagery from a focal plane array (102), and decomposing (404) the video imagery into independently-moving coordinate transformations corresponding to clutter patterns that are subimages of the video imagery. The method also includes removing (406) the subimages from an image of the video imagery to produce a clutter-suppressed version of the image, and rendering (416) the clutter-suppressed version of the image. Removing (406) the subimages from the image includes generating (408) respective zeroth-order corrections for the subimages, and subtracting (410) the respective zeroth-order corrections from the image. And removing (406) the subimages includes recursively generating (412) respective first-order corrections for the subimages from the respective zeroth-order corrections, and subtracting (414) the respective first-order corrections from the image. The respective zeroth-order corrections are generated and subtracted from the image before the respective first-order corrections are generated and subtracted from the image.