Particulate Contamination Classification via Multi-Angle Illumination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for classifying particulate contaminants on surfaces require complex extraction processes and laboratory analysis, making it difficult to classify materials directly in the production process.
Innovation Solution
A method involving the illumination of contaminants with electromagnetic radiation from different directions, capturing images under varying conditions, and calculating a measure of similarity between these images to classify the contaminants into different material classes based on their surface properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the established standard method (VDA 19 or ISO 16232) is used to classify particulate contaminants, then the classification accuracy is improved, but the process complexity and time consumption increase due to required extraction and laboratory analysis
Solution Approach 1:
The invention extracts and analyzes only the essential optical properties (gloss characteristics) of particles directly on the surface, eliminating the need to extract particles from the rinsing liquid and perform complex laboratory analysis. The method focuses solely on capturing and evaluating reflectance images to determine particle material composition.
Solution Approach 2:
The invention creates optical copies (images) of particles under different illumination conditions and analyzes these copies to determine material properties. By capturing reflectance characteristics from multiple angles and comparing them, the system identifies particle material without physical extraction or laboratory processing.
2Reliability
If the established standard method is used, then the classification reliability is improved, but the productivity decreases due to time-consuming extraction and laboratory analysis
Solution Approach 1:
The invention performs classification actions during the rinsing process itself, before particles need to be extracted and sent to the laboratory. By capturing optical images of particles on the surface during production, the system enables real-time or near-real-time classification, eliminating the waiting time associated with laboratory analysis.
Solution Approach 2:
The invention replaces the mechanical extraction and laboratory microscopy process with an optical imaging and image processing system. Instead of physically removing particles and analyzing them under a microscope, the system uses cameras to capture reflectance characteristics and algorithms to automatically classify particles, significantly increasing processing speed.
3Speed
If optical systems are used to detect particulate contaminants, then the detection speed is improved, but the ability to classify particle types is lost
Solution Approach 1:
The invention analyzes local optical properties (gloss characteristics) of individual particles by capturing images under different illumination angles. By examining the reflectance patterns specific to each particle's surface properties, the system determines material composition while maintaining high detection speed through automated image processing.
Solution Approach 2:
The system changes illumination parameters (angles and directions) to extract different optical characteristics from particles. By capturing images under varying illumination conditions and analyzing how reflectance changes with angle, the system identifies material properties that distinguish different particle types while maintaining rapid detection capability.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables direct classification of particulate contaminants on the surface, eliminating the need for laboratory analysis and allowing for real-time material identification in the production process, effectively distinguishing between materials with high and low gloss based on image similarity.
Implementation Method 1
particulate contaminants with different surface properties have different gloss. Gloss is understood to mean the ability of a surface to reflect electromagnetic radiation following the law of reflection
Data Source
Figure 1~2
Figure 3
AI summary
The present invention relates to a method for classifying a particulate contaminant (2) on a surface (1) of a test specimen (7) comprising the steps of: illuminating the contaminant (2) with electromagnetic radiation with a first illumination direction (9) relative to the contaminant (2), capturing a first image of the contaminant (2) illuminated by the radiation from the first illumination direction (9) with a first observation direction (8) relative to the contaminant (2), illuminating the contaminant (2) with electromagnetic radiation with a second illumination direction (10) relative to the contaminant (2), capturing a second image of the contaminant (2) illuminated by the radiation from the second illumination direction (10) with a second observation direction relative to the contaminant (2),where a) either the first illumination direction (9) and the second illumination direction (10) are different and the first observation direction (8) and the second observation direction are the same, or b) the first illumination direction (9) and the second illumination direction (10) are the same and the first observation direction (8) and the second observation direction are different, calculating a measure of similarity between the first image and the second image and classifying the contamination (2) into at least a first and a second class based on the calculated measure of similarity.