Altered Currency Detection via Grayscale Image Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for identifying altered paper money are inaccurate due to the simplification of grayscale values to two levels, making it difficult to detect subtle differences in adjacent pixels, which can lead to false negatives in identifying altered currency.

Innovation Solution

A paper money identification method that divides the grayscale image into multiple areas, calculates feature values for each area, and compares these values with a template to determine if they match, with a threshold-based system to count unmatched areas and determine if the paper money is altered.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If grayscale values are simplified to two levels (0 or 255) through binarization, then processing speed and computational simplicity are improved, but measurement precision deteriorates because adjacent pixels with minor grayscale value differences cannot be detected

Engineering Contradiction:
Improveprocessing speedVSAvoidgrayscale value detection precision
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent divides the grayscale image into multiple small blocks (e.g., 10x10 pixel blocks) and processes each block independently. Within each block, it calculates grayscale statistics (mean, standard deviation) and compares adjacent blocks to detect spliced seams. This segmentation allows the system to maintain full grayscale precision (256 levels) while reducing the computational complexity of pixel-by-pixel comparison across the entire image.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Instead of processing all pixels with full precision, the patent applies partial action by only performing detailed grayscale analysis on specific regions of interest (blocks near detected edges or areas with high variance). This reduces the overall computational load while maintaining detection precision where it matters most.

Inventive Principle:
Principle #16Partial or excessive action

2Reliability

If all pixels in the binarized image are compared dot-to-dot to detect spliced seams, then detection thoroughness is improved, but loss of time increases due to the large number of comparisons required

Engineering Contradiction:
Improvedetection thoroughnessVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the image into blocks and only performs detailed comparisons between adjacent blocks rather than all pixels. This reduces the number of comparisons from millions (in a full-resolution image) to thousands (between blocks), dramatically reducing processing time while maintaining detection reliability through block-level statistical analysis.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary processing by dividing the image into blocks and calculating grayscale statistics for each block before performing comparisons. This preliminary organization of data structures and pre-computation of block characteristics reduces the computational complexity of the subsequent comparison stage.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If the entire grayscale image is processed to calculate feature values for each area, then identification accuracy is improved, but device complexity increases due to the multi-step processing required

Engineering Contradiction:
Improveidentification accuracyVSAvoidprocessing system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the complex task of image analysis into segmented steps: block division, grayscale statistic calculation, adjacent block comparison, and spliced seam detection. Each step handles a specific aspect of the problem, making the overall system more manageable and implementable despite the increased processing requirements for high accuracy.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP2889846B1Paper money identification method and device
Publication Date: 2019.03.06 SHANDONG NEW BEIYANG INFORMATION TECH CO LTD
  • EP2889846B1 patent drawingFigure 1a~1b
  • EP2889846B1 patent drawingFigure 2a~2b
  • EP2889846B1 patent drawingFigure 3~4

AI summary

The invention discloses a paper money identification method and device. The paper money identification method comprises: obtaining entire grayscale image data of the paper money to be tested; dividing the entire grayscale image of the paper money to be tested into a plurality of areas, and calculating a feature value of the image data of each area respectively; comparing the feature value of each area with the feature value of a corresponding area in a paper money template to judge whether the two compared areas are matched, wherein, the two compared areas are judged to be unmatched if a difference between the feature values of the two compared areas is greater than a first preset threshold, otherwise the two compared areas are judged to be matched; counting the number of unmatched areas of the paper money to be tested; judging whether the number of the unmatched areas of the paper money to be tested is greater than a second preset threshold; and determining that the paper money to be tested is the altered money when the number of the unmatched areas of the paper money to be tested is judged to be greater than the second preset threshold. By the invention, the accuracy of identifying the altered money is improved.