Sparse Adaptive Filter for Hyperspectral Target Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional hyperspectral image processing systems face computational intensity in detecting targets due to processing large amounts of data, with only a fraction being useful, and existing band selection methods are not adaptive to specific target-scene combinations, leading to inefficient processing and potential local minima in search results.
Innovation Solution
A filtering engine that dynamically and adaptively selects a subset of hyperspectral wavebands using a layered approach, including Ranking, Leave One Out, and Variable Band Search techniques, based on signal-to-clutter ratio, to reduce computational complexity and identify a 'good enough' set of bands for target detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all hyperspectral wavebands are processed to detect targets, then detection accuracy is improved, but computational complexity increases significantly
Solution Approach 1:
The patent extracts and processes only a selected subset of hyperspectral wavebands that contain useful information for target detection, rather than processing all available wavebands. This is achieved through iterative filtering that identifies and removes redundant bands while preserving those critical for detection accuracy.
Solution Approach 2:
The patent segments the full hyperspectral dataset into a smaller, manageable subset of informative wavebands. By dividing the complete spectrum into selected discrete bands, the system reduces computational load while maintaining detection performance.
2Productivity
If a fixed band selection method is used, then processing efficiency is improved, but adaptability to different target-scene combinations deteriorates
Solution Approach 1:
The patent implements a dynamic band selection process that adapts to different target-scene combinations. The filtering engine iteratively adjusts which wavebands are selected based on the specific characteristics of each detection scenario, allowing the system to optimize performance for varying conditions rather than using a static band set.
Solution Approach 2:
The patent changes the parameters of band selection based on the specific target and scene being analyzed. By modifying which wavebands are included in the processing subset according to the detection requirements, the system achieves both efficiency and adaptability.
3Measurement precision
If exhaustive band selection search is performed, then band selection accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs a partial search through the space of possible band combinations rather than an exhaustive enumeration. The iterative filtering process evaluates and selects bands based on their contribution to detection accuracy, achieving good band selection without examining all possible combinations.
Solution Approach 2:
The patent performs preliminary filtering to identify and eliminate obviously redundant wavebands before final selection. This preliminary action reduces the search space and allows the system to achieve accurate band selection more efficiently.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The disclosure provides a filtering engine for selecting a subset of hyperspectral imaging wavebands having information useful for detecting a target in a scene. Selecting these wavebands, called "sparse bands," is an iterative process. One or more search techniques of varying computational complexity are used in the process. The techniques rely on various selection criteria, including a signal to clutter ratio that measures the "goodness" of band selection. A convenient example of the filtering engine uses several of the techniques together in a layered approach. In this novel approach, simpler computational techniques are applied, initially, to reduce a number of bands. More computationally intensive techniques then search the reduced band space. Accordingly, the filtering engine efficiently selects a set of sparse bands tailored for each target and each scene, and maintains some of the detection capability provided with a full set of wavebands.