Automated Defect Categorization for Banknote Fitness Assessment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Automated defect categorization in media items, such as banknotes, deposited in self-service terminals is challenging due to the absence of human inspection, making it difficult for currency issuing authorities to identify and return unfit currency.

Innovation Solution

A method and system that involves receiving a binarized image of the media item, identifying blobs with potential defect intensity, categorizing them as tears, missing portions, corner folds, or holes based on size and location, and accessing a defect profile configuration file to determine the media item's fitness for continued circulation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If automated defect detection is implemented in self-service terminals, then productivity is improved, but measurement precision deteriorates due to absence of human inspection

Engineering Contradiction:
Improvedefect detection efficiencyVSAvoiddefect categorization accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent replaces human mechanical inspection with an automated image processing system that uses computer vision algorithms to detect, segment, and categorize defects. The system processes images through multiple stages including binarization, blob detection, and classification algorithms to automatically identify defect types such as tears, stains, and missing portions, thereby maintaining high productivity while achieving reliable measurement precision through computational methods.

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

Solution Approach 2:

The system creates digital copies (images) of the media items for analysis, allowing multiple processing stages and algorithms to evaluate the same defect without physically handling the original item. This copying approach enables repeated analysis and verification, improving categorization accuracy while maintaining automated high-speed processing.

Inventive Principle:
Principle #26Copying

2Reliability

If multiple defect categories are identified, then reliability is improved, but device complexity increases

Engineering Contradiction:
Improvecurrency fitness assessment accuracyVSAvoidcategorization system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the defect detection process into distinct stages: image acquisition, binarization, blob detection, defect type classification, and fitness assessment. Each stage handles a specific aspect of analysis, with dedicated algorithms for different defect types (tears, stains, missing portions, corner damage). This segmentation improves reliability by ensuring thorough evaluation while managing complexity through modular processing steps.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system transitions from analyzing simple image pixels to identifying two-dimensional blobs with spatial properties (area, perimeter, shape factors). By adding dimensional analysis of defect geometry and position, the system achieves more reliable categorization without proportionally increasing overall system complexity, as the additional analysis is performed through algorithmic processing rather than physical components.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP2660787B1Defect Categorisation in a digital image
Publication Date: 2015.10.14 NCR VOYIX CORP
  • EP2660787B1 patent drawingFigure 1~6
  • EP2660787B1 patent drawingFigure 2A
  • EP2660787B1 patent drawingFigure 2B

AI summary

A method of categorising defects in a media item. The method comprising the steps of: receiving a binarised image of the media item (step 112), where the binarised image comprises a plurality of pixels, each pixel having either a potential defect intensity or a non-defect intensity; and identifying one or more blobs comprising contiguous pixels each having a potential defect intensity (step 114). For each identified blob, the method involves comparing a size of the blob with a damage threshold (step 116); ignoring the blob if the blob size is smaller than the damage threshold (step 118); and for each identified blob having a size exceeding or equalling the damage threshold, categorising the identified blob.