Reconstructing Reflection Coefficients via L1-Norm Minimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current imaging methods for determining the internal structure of specimens often smooth out discrete structures, leading to inaccurate representations of the internal structure due to assumptions about smoothness, and are sensitive to errors in measurement series, which can destabilize the reconstruction process.

Innovation Solution

A method for reconstructing the spatial distribution of reflection coefficients by parameterizing the specimen into volume elements, creating route matrices for different measurement configurations, and merging measurement series and matrices to solve for reflection coefficients, allowing for sparse sampling that violates Nyquist's theorem, enabling sharper structure representation and robust error handling.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Stability of the object's composition

If smoothness assumption is applied according to Occam's principle, then the solution stability is improved, but the sharpness of discrete structures is lost

Engineering Contradiction:
Improvesolution stabilityVSAvoidsharpness of discrete structures
Core Design Contradiction:
Stability of the object's compositionVSManufacturing precision

Solution Approach 1:

The patent changes the boundary condition from a smoothness constraint to a sparsity constraint on the reflection coefficient vector. By minimizing the L1-norm of the reflection coefficient vector instead of assuming smoothness, the method enables sharp discrete structures to be reconstructed while maintaining solution stability through the sparsity prior.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent applies local quality by allowing different regions of the specimen to have different properties: most regions have zero reflection coefficient (homogeneous) while specific localized regions have non-zero reflection coefficients representing discrete structures. This sparse representation captures sharp local features without requiring global smoothness.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If complex filtering is applied to measurement series, then the measurement precision is improved, but the device complexity increases

Engineering Contradiction:
Improvemeasurement precisionVSAvoidcomplexity of filtering equipment
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces complex physical filtering equipment with computational signal processing. By using L1-norm minimization and sparse reconstruction algorithms, the method achieves accurate measurement results through software-based error handling and signal processing rather than requiring complex hardware filtering systems.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces an intermediary computational model (sparse reconstruction algorithm) that mediates between the raw measurements and the final image. This computational intermediary handles errors and noise in the measurement series without requiring complex physical filtering, simplifying the overall system while maintaining precision.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Manufacturing precision

If sufficient measurement series are recorded to ensure accuracy, then the reconstruction quality is improved, but the measurement time increases

Engineering Contradiction:
Improvereconstruction qualityVSAvoidmeasurement time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent applies partial action by using fewer measurement series than traditionally required, compensated by the sparsity constraint. The L1-norm minimization allows accurate reconstruction with incomplete or reduced measurement data, reducing measurement time while maintaining or improving reconstruction quality through the power of sparse optimization.

Inventive Principle:
Principle #16Partial or excessive action

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 allows for the reconstruction of sharply defined internal structures with simpler and cost-effective equipment, reducing the need for complex filtering and achieving accurate three-dimensional distribution of reflection coefficients, even when measurements violate Shannon's sampling theorem.

Implementation Method 1

wave-mechanical interactions, such as reflections or scattering, of the sample body can be measured

Methodology Applied
Scientific EffectReflection: Reflection

Implementation Method 2

wave-mechanical interactions, such as reflections or scattering, of the sample body can be measured

Methodology Applied
Scientific EffectScattering: Scattering

Data Source

PatentEP2539870B1Method for reconstructing the internal structure of a sample body by means of reflection and scattering signals
Publication Date: 2014.08.06 KARLSRUHER INST FUR TECH
  • EP2539870B1 patent drawingFigure 1
  • EP2539870B1 patent drawingFigure 2~3
  • EP2539870B1 patent drawingFigure 4a~4f

AI summary

One aspect of the invention relates to a method for reconstructing the spatial distribution of a reflection coefficient for waves in a sample body, comprising the following steps: parameterizing the sample body by means of N volume elements; determining M different measurement configurations of an emitting device and an associated receiving device; defining the number T of measurement points; setting up M path matrices G[1], wherein the possible paths of reflected or scattered waves are encoded in the i-th path matrix; registering M series of measurements, wherein an excitation signal a[i] is fed to the emitting device in the i-th series of measurements, and the associated receiving device (5) registers a series of measurements x[i]; calculating a predictive differential vector Δx = x- G-s [n], which contains the difference between the elements of the M registered series of measurements x[i] and the predictable series of measurements G[i] tj-s[n] j; minimizing a norm of the predictive differential vector || Δx ||; and providing the reflection coefficients s[n]. The invention also relates to an apparatus for carrying out the method.