Particle Image Analysis Background Region Abnormality Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional flow type particle image analysis apparatuses are unable to detect all abnormalities caused by human error, sample abnormalities, and apparatus abnormalities, leading to delayed detection of measurement errors and incorrect results, which reduces inspection efficiency.
Innovation Solution
The apparatus simultaneously analyzes the background region of captured images during normal particle image analysis to determine abnormalities such as sample shortages, sheath liquid shortages, and flow path system abnormalities, using image processing to identify peaks and ratios that indicate abnormal conditions, allowing for immediate detection and alerting operators without adding complex mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional flow type particle image analysis apparatuses are used with separate abnormality determination processes, then the apparatus configuration remains simple, but abnormality detection is delayed and inspection efficiency is reduced
Solution Approach 1:
The patent merges the abnormality determination process with the normal particle image analysis process by using the same captured images for both purposes. The background region analysis is integrated into the existing image processing workflow, allowing simultaneous detection of particles and abnormalities without requiring separate measurement processes or additional hardware.
Solution Approach 2:
The patent performs abnormality detection during the normal measurement process rather than after completion. By analyzing the background region of captured images in real-time alongside particle analysis, the system detects abnormalities immediately when they occur, enabling prompt operator response and avoiding delayed detection after all measurements are completed.
2Reliability
If sensor-based methods are used to detect liquid surfaces at sample vessels or reagent locations, then sample and reagent shortages can be detected, but the apparatus configuration becomes more complicated and flow path abnormalities cannot be detected
Solution Approach 1:
The patent replaces physical sensor-based detection methods with image processing-based detection. Instead of using sensors to detect liquid surfaces at sample vessels or reagent locations, the system uses the existing imaging system to capture images and analyzes the background region for characteristics indicating abnormalities such as sample shortages, sheath liquid shortages, or flow path issues. This substitution eliminates the need for additional sensors while expanding detection capability to include flow path abnormalities.
Solution Approach 2:
The patent makes the imaging system serve multiple functions: it detects particles for analysis, detects abnormalities in the measurement process, and identifies issues throughout the entire flow path system. By analyzing the background region of captured images, the same imaging apparatus performs what would otherwise require separate sensor systems, achieving multi-functionality without increasing device complexity.
3Reliability
If background region analysis is conducted simultaneously with normal particle image analysis, then all types of abnormalities can be detected without delaying measurement, but the image processing load increases
Solution Approach 1:
The patent segments the image analysis into two independent components: particle detection and abnormality determination. The abnormality determination focuses specifically on analyzing the background region characteristics (such as color distribution, texture, or patterns) rather than processing the entire image at full detail. This segmentation allows parallel processing of particle and abnormality analysis while minimizing the computational load for each task.
Data Source
Figure 1
Figure 2
Figure 3(a)~3(c)
AI summary
A conventional flow type particle analysis apparatus could not detect both a human error, such as leaving behind a sample, a shortage of the sample, or a shortage of a reagent, and an abnormality of a flow path system by an inexpensive and simple method. A particle analysis apparatus according to the present invention includes means that perform processes upon image processing of images (110) acquired in a measurement of the sample, simultaneously with a normal image processing for classifying target particles. The means included are: image processing means (110a) for calculating information of RGB density distributions of each whole image; abnormal state determination processing means (110c) for determining whether or not the acquired images are in an abnormal state according to tendencies of the RGB density distributions; and an abnormality judgment process means for making final determination of the existence of an abnormality by calculating an appearance frequency of abnormal images after all measurements for the one sample is completed. These means allow diagnosis of abnormalities to be conducted simultaneously with normal analysis without changing the configuration of the conventional apparatus.