Hardness Tester Image Processing for Accurate Vertex Extraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional hardness testing methods face reliability issues due to dependence on noise reduction accuracy and misrecognition of impression shapes, especially when dealing with multiple samples or scratches parallel to the impression region, leading to inaccurate results.

Innovation Solution

A hardness tester and method that extracts impression regions and vertices using binarized image data with reduction and expansion processing, followed by distance conversion, allowing for accurate calculation of hardness without relying on noise reduction accuracy and enabling the detection of multiple samples.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If conventional techniques use noise reduction processing on binary images to determine impression boundaries, then the processing can be simplified, but the reliability of vertex position determination deteriorates due to dependence on noise reduction accuracy

Engineering Contradiction:
Improveimage processing complexityVSAvoidvertex position determination reliability
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent applies preliminary action by performing reduction processing and expansion processing on the binary image before boundary detection. This pre-processing removes noise and fills gaps in the impression boundaries, creating a cleaner image that leads to more reliable vertex position determination without requiring complex noise reduction algorithms later in the process.

Inventive Principle:
Principle #10Preliminary action

2Adaptability or versatility

If conventional techniques estimate impression regions using approximated multidimensional curves, then the method can handle complex shapes, but the measurement precision of vertex positions deteriorates

Engineering Contradiction:
Improveimpression shape handling capabilityVSAvoidvertex position measurement precision
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent applies inversion by reversing the conventional approach: instead of estimating curves and finding their intersections, it directly detects boundary points from the pre-processed binary image and uses these actual boundary points to determine vertex positions. This inversion from estimation to direct detection significantly improves measurement precision while maintaining the ability to handle complex impression shapes.

Inventive Principle:
Principle #13The other way round (Inversion)

3Device complexity

If conventional techniques assume only one impression in the captured image, then the processing can be simplified, but the adaptability to test multiple samples deteriorates

Engineering Contradiction:
Improveimage processing complexityVSAvoidmultiple samples detection capability
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent applies universality by designing an image processing system that can handle both single and multiple impressions using the same core algorithms. The boundary detection and vertex determination methods work universally for any number of impressions in the image, allowing the system to adapt to different test scenarios (single sample or multiple samples) without requiring different processing approaches.

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

Data Source

PatentEP2570793B1Hardness tester and hardness test method
Publication Date: 2019.04.17 MITUTOYO CORP
  • EP2570793B1 patent drawingFigure 1
  • EP2570793B1 patent drawingFigure 2
  • EP2570793B1 patent drawingFigure 3

AI summary

A hardness tester (100) includes an image capture control means (61) for obtaining an image data of the surface of the sample by controlling and making an image capture means (12) capture an image of the surface; an impression region extracting means (61) that binarizes the image data, applies reduction/expansion processing to the binarized image data, applies a distance conversion processing to the reduction/expansion processed image data, and extracts a closed region corresponding to a contour of the indenter by using the distance-converted image data; an impression vertex extracting means (61) that estimates the vertex for measuring the impression based on a profile of the closed region and extracts a point in the binarized image data, as the vertex for measuring the impression, that agrees with a predetermined condition; and a hardness calculating means (61) for calculating hardness of the sample based on the vertexes for measuring the impression.