Sample Surface State Detection Using Bubble Size Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing automatic analysis systems face inefficiencies in sample analysis due to the detection of air bubbles, as current methods either overlook bubble size or struggle with varying sample conditions, leading to inaccurate analysis and reduced efficiency.

Innovation Solution

An image processing system that determines the state of a sample surface by considering the position and size of air bubbles relative to a detection range, using machine learning and convolutional neural networks to accurately classify the presence and size of air bubbles, thereby optimizing sample acquisition and analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If air bubble detection is performed using edge component analysis or color difference methods, then air bubbles can be detected, but small air bubbles that do not affect analysis are also detected, reducing analysis efficiency

Engineering Contradiction:
Improveair bubble detection accuracyVSAvoidanalysis efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies local quality by defining a specific detection range centered on the container opening portion rather than analyzing the entire image. By setting the detection range to cover the area where the dispensing probe contacts the sample surface, the system focuses computational resources on the relevant region, detecting only air bubbles that would interfere with sample acquisition while ignoring smaller bubbles elsewhere that do not affect analysis.

Inventive Principle:
Principle #3Local quality

2Device complexity

If air bubble detection is performed using conventional methods, then detection can be done, but large air bubbles covering the detection range generate edges outside the detection range, making them undetectable

Engineering Contradiction:
Improvedetection method simplicityVSAvoidlarge air bubble detection accuracy
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent transitions from edge-based detection to area-based detection by calculating the area ratio of white pixels within the detection range. This dimensional change from detecting boundaries (1D edges) to measuring filled areas (2D regions) enables reliable detection of large air bubbles that may not produce detectable edges within the constrained detection range, as their substantial area occupies enough of the detection region to be identified.

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

3Quantity of substance

If liquid level detection is used to control dispensing probe immersion, then sample delivery amount can be controlled, but air bubbles on the sample surface cause false detection, preventing sufficient sample acquisition

Engineering Contradiction:
Improvesample delivery amount controlVSAvoidliquid level detection accuracy
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent introduces an intermediary air bubble detection step between image capture and liquid level determination. By first detecting and flagging the presence of air bubbles in the detection range before proceeding with liquid level measurement, the system prevents false liquid level detection caused by air bubbles, ensuring that sample acquisition is only inhibited when truly necessary.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3663767B1Apparatus, method for determing state of sample, and analysis system
Publication Date: 2023.07.26 HITACHI HIGH TECH CORP
  • EP3663767B1 patent drawingFigure 1
  • EP3663767B1 patent drawingFigure 2~3A
  • EP3663767B1 patent drawingFigure 3B

AI summary

A state of a sample surface is accurately determined without lowering analysis efficiency. There is provided an apparatus for determining a state of a sample to be analyzed contained in a container, in which the apparatus acquires an image of the sample, analyzes a position and a size of an object to be detected with respect to a detection range set in the image by using the image of the sample, and determines the state of the sample based on a result of the analysis.