Object Hashing via Fourier Transform Coefficient Selection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional traceability methods, such as barcodes and RFID, are inadequate for tracking individual components throughout their entire production chain due to vulnerabilities to counterfeiting, high costs, and inability to track components from the first production step, and they are not robust against surface alterations during manufacturing processes.

Innovation Solution

A method using a multi-dimensional transform, specifically a 2D discrete Fourier transform, to obtain a hash value from an object's surface image, selecting specific transform coefficients to create a unique identifier that is robust and efficient for tracking components, reducing computational costs and storage requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional traceability methods (barcode, RFID, DMC) are used to track components, then individual identification is achieved, but the system becomes vulnerable to counterfeiting and requires additional labeling processes that increase costs and alter the original part

Engineering Contradiction:
Improveanti-counterfeiting capabilityVSAvoidlabeling process complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The object's own surface serves as the identification carrier, eliminating the need for external labels or tags. The inherent surface patterns naturally provide unique identification without requiring additional components or processes

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The identification function is extracted from separate labeling components and integrated directly into the object's surface characteristics, using the surface itself as the identification medium

Inventive Principle:
Principle #2Taking out (Extraction)

2Ease of operation

If labels or RFID tags are attached to parts for traceability, then individual tracking is enabled, but running costs increase and the original part is altered

Engineering Contradiction:
Improvetraceability functionalityVSAvoidmanufacturing cost
Core Design Contradiction:
Ease of operationVSEase of manufacture

Solution Approach 1:

The object's surface provides the identification function inherently, eliminating the need for separate labeling components and their associated manufacturing and attachment costs

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The surface serves dual purposes: both as the functional surface of the part and as the identification medium, combining manufacturing and traceability functions

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Measurement precision

If surface patterns are captured for identification, then unique object characterization is achieved, but the system becomes sensitive to surface alterations during manufacturing

Engineering Contradiction:
Improveobject characterization accuracyVSAvoidrobustness against surface alterations
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The identification data is extracted from specific frequency components of the surface pattern that are less susceptible to manufacturing alterations, isolating the stable characteristic features

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The identification approach transitions from using raw surface images to using frequency domain representations (transform coefficients), which provide more stable and reliable identification characteristics

Inventive Principle:
Principle #35Parameter changes

4Loss of information

If full surface images are used for identification, then comprehensive object information is obtained, but storage and computational requirements increase

Engineering Contradiction:
Improveidentification information completenessVSAvoiddata storage requirement
Core Design Contradiction:
Loss of informationVSQuantity of substance

Solution Approach 1:

Only the essential identification information is extracted from the full surface image through frequency transformation and selective coefficient extraction, discarding redundant data

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The surface pattern information is segmented into frequency components, and only the relevant coefficients that provide unique identification are retained for storage and comparison

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3840289A1Apparatus for obtaining a hash value associated with an object, object identifier, methods and computer program
Publication Date: 2021.06.23 DETAGTO GMBH
  • EP3840289A1 patent drawingFigure 1~2
  • EP3840289A1 patent drawingFigure 3a~3b
  • EP3840289A1 patent drawingFigure 4~5

AI summary

This invention is related to an apparatus (100; 150) for obtaining a hash value (100a) associated with an object (110). The apparatus (100; 150) is configured to apply a multi-dimensional transform to an image (120a) of the object (110) or of a surface of the object (110), in order to obtain a multi-dimensional array (600) of transform coefficients. Furthermore, the apparatus (100; 150) is configured to obtain the hash value (100a) using transform coefficients (1110), for which a first of the transform coefficient indices is smaller than a first predetermined value but for which a second of the transform coefficient indices is equal to or larger than a second predetermined value. In addition, the apparatus (100; 150) is configured to obtain the hash value (100a) using transform coefficients (1110) for which the second of the transform coefficient indices is smaller than the second predetermined value but for which the first of the transform coefficient indices is equal to or larger than the first predetermined value. Moreover, the apparatus (100; 150) is configured to obtain the hash value (100a) while leaving transform coefficients (1110) for which the first transform coefficient index is smaller than the first predetermined value and for which the second transform coefficient index is smaller than the second predetermined value unconsidered in the determination of the hash value (100a).