Spatio-Spectral Edge Detection in Hyperspectral Imagery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing edge detection algorithms for multispectral and hyperspectral images face challenges in detecting isoluminant edges, as they often require high computational resources and struggle with scalability, missing edges in some image planes, and lack standard approaches for fusing information across planes.
Innovation Solution
The development of the Spectral Ratio Contrast (SRC) and Adaptive Spectral Ratio Contrast (ASRC) algorithms, which utilize spectral-bands ratios and material classification to define a sparse, three-dimensional mask that fuses spectral and spatial information for edge detection, avoiding derivative-based methods and adapting sensitivity to material changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard gradient-based operators are used for edge detection in spectral images, then intensity edges can be detected, but isoluminant edges cannot be detected effectively
Solution Approach 1:
The patent transitions from 2D spatial gradient operations to 3D spatio-spectral operations by incorporating spectral band information. The spatio-spectral mask operates in three dimensions (x, y, and spectral bands), enabling detection of edges that manifest differently across spectral dimensions rather than relying solely on spatial intensity gradients.
Solution Approach 2:
The patent changes the detection parameters by using spectral ratios between different bands instead of absolute intensity values. This parameter transformation allows the detection of isoluminant edges where intensity remains constant but spectral composition changes, overcoming the limitation of gradient-based operators.
2Ease of manufacture
If multiple image planes are processed separately using standard edge detectors, then processing simplicity is maintained, but information fusion across planes becomes ad hoc and edges may be missed
Solution Approach 1:
The patent merges the processing of multiple image planes into a unified spatio-spectral operation. Instead of separately processing each spectral band and then combining results, the spatio-spectral mask applies a single integrated operation across all bands simultaneously, ensuring consistent and complete edge detection.
Solution Approach 2:
The spatio-spectral mask serves multiple functions: it detects edges in spatial dimensions, captures spectral variations, and fuses information across all spectral bands in a single operation. This multi-functional approach replaces the need for separate processing and ad hoc fusion of individual band results.
3Measurement precision
If the Multicolor Gradient (MCG) algorithm is used for joint spatio-spectral edge detection, then isoluminant edges can be detected, but computational complexity becomes prohibitively high
Solution Approach 1:
The patent extracts and utilizes only the essential spectral information needed for edge detection by focusing on spectral ratios between specific bands. This selective extraction avoids the computationally intensive processing of all spectral bands required by MCG, reducing complexity while maintaining detection capability.
Solution Approach 2:
The patent applies partial action by using a simplified spatio-spectral mask that processes a subset of spectral information rather than the complete spectral data. This partial processing approach achieves sufficient edge detection performance with significantly reduced computational burden compared to the full MCG algorithm.
4Productivity
If spectral bands are reduced to lower computational complexity, then processing efficiency improves, but detection accuracy may be compromised
Solution Approach 1:
The patent transforms the spectral data by computing ratios between bands, which concentrates the relevant edge information into fewer derived parameters. This parameter transformation maintains detection accuracy by preserving spectral contrast information while reducing the data volume requiring processing.
Data Source
AI summary
Apparatus, systems, and methods fusing material classification with spatio-spectral edge detection in spectral imagery can be used in a variety of applications. In various embodiments, a classifier can be applied to neighboring pixels in data for an image to determine, based on material changes, if the neighboring pixels are correlated to two different materials with respect to a candidate location for an edge. Results of the classification can be used with a spatio-spectral mask to accept or reject the candidate location as an edge. Additional apparatus, systems, and methods are disclosed.