Hyperspectral Target Identification Using Two-Stage Library Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hyperspectral imagery analysis faces challenges in material identification due to insufficient specificity in library spectra and the limitations of statistical regression methods, such as high false alarm rates and difficulty in distinguishing between similar materials, particularly in distinguishing target materials from confusers and identifying materials at multiple levels of resolution.
Innovation Solution
A two-stage identification system integrating F-Test and model averaging approaches, combined with a multi-tier target library and hierarchical identification, which provides detailed information by computing probabilities at multiple levels of specificity and using separate libraries for detection and identification to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a single library is used for both target detection and material identification, then the system is simpler to operate, but the detection accuracy and identification specificity are reduced
Solution Approach 1:
The patent divides the single library into two separate libraries: a target detection library containing spectra of materials of interest, and a material identification library containing spectra of potential confuser materials. This segmentation allows each library to be optimized for its specific purpose, improving both detection sensitivity and identification accuracy while maintaining operational simplicity through automated management.
2Productivity
If statistical regression methods are used for material identification, then the process is computationally efficient, but false alarm rates increase and difficulty in distinguishing similar materials persists
Solution Approach 1:
The patent introduces an intermediary comparison process between the target detection library and material identification library. After initial target detection, the system mediates by comparing detected targets against confuser spectra in the identification library, allowing computationally efficient processing while improving reliability through an additional verification step that reduces false alarms.
3Adaptability or versatility
If library spectra lack sufficient specificity, then the system is more adaptable to various materials, but the ability to distinguish between similar materials deteriorates
Solution Approach 1:
The patent applies local quality by creating specialized libraries with different spectral resolutions and specificities tailored to their purposes. The target detection library uses broader spectral characteristics for high adaptability, while the material identification library incorporates more specific spectral features to distinguish between similar materials, allowing each component to have optimized properties for its function.
Data Source
AI summary
Embodiments of a system, method and apparatus incorporate two-stage identification of targets in hyperspectral image analysis. In various embodiments, unmixing is employed that integrates F-test and model averaging approaches. Further, a multi-tier target library process provides an improvement in the spectra that can be used to detect target materials and spectra that can be used for unmixing in identification. Additionally, the hierarchical identification of the present disclosure combines probabilities from model averaging to generate target identifications simultaneously at multiple levels of specificity.