Spectroscopic Camera Attached Substance Detection Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing attached substance determination devices struggle to accurately detect substances on products with similar base colors or transparent substances, as human visual methods are inadequate and fail to distinguish between base and attached substances.
Innovation Solution
A method using a spectroscopic camera and machine learning to generate a determination model based on spectroscopic images of samples with and without attached substances, allowing for accurate detection of attached substances by analyzing differences in spectroscopic spectra across various wavelengths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human visual characteristic-based determination is used, then the determination method is simple and fast, but it cannot detect attached substances with similar base colors or transparent substances
Solution Approach 1:
The patent replaces human visual inspection with a spectroscopic imaging system that captures spectral data across multiple wavelengths. The spectroscopic camera captures images at different wavelengths (e.g., blue, cyan, green, yellow, orange, red), and a determination model processes these spectral data to detect attached substances. This substitution enables detection of substances invisible to human eyes, such as transparent oils or color-matched contaminants, by analyzing spectral characteristics rather than visual appearance.
Solution Approach 2:
The patent changes the detection parameter from visible color (human visual characteristic) to spectroscopic wavelength data. By capturing and analyzing multiple wavelength bands, the system can detect subtle spectral differences caused by attached substances even when they have similar colors or are transparent. The determination model processes spectral data across different wavelengths to identify patterns that indicate the presence of attached substances, transforming the detection approach from qualitative visual assessment to quantitative spectral analysis.
2Measurement precision
If spectroscopic imaging with multiple wavelength samples is used, then detection accuracy for all types of attached substances is improved, but the number of required samples and processing complexity increases
Solution Approach 1:
The patent creates a universal determination model that can detect various types of attached substances (colored, transparent, color-matched) using the same multi-wavelength spectroscopic imaging approach. The model is trained on diverse sample data representing different substance types and base colors, enabling it to generalize and accurately detect any attached substance regardless of its specific properties. This multi-functional capability allows a single system to handle diverse inspection scenarios without requiring separate specialized methods for each substance type.
Solution Approach 2:
The patent performs preliminary actions by pre-acquiring spectroscopic images of multiple wavelength samples for both samples with and without attached substances during the training phase. This preliminary data collection and model training enables the determination model to learn the spectral characteristics of different substances and bases in advance. During actual inspection, the model can quickly compare unknown samples against this pre-learned knowledge base, reducing the need for real-time complex analysis and enabling rapid accurate detection.
3Measurement precision
If conventional image processing is used, then processing speed is fast, but it cannot distinguish between base color and attached substance color
Solution Approach 1:
The patent segments the spectral data into multiple wavelength bands (blue, cyan, green, yellow, orange, red) and analyzes each band separately to identify characteristic absorption patterns. By segmenting the spectral information, the system can distinguish between the spectral signature of the base material and the spectral signature of the attached substance, even when they have similar visible colors. The determination model processes these segmented spectral data to identify subtle differences that conventional single-color image processing cannot detect.
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 high-accuracy detection of attached substances, including transparent and color-matched substances, by analyzing spectroscopic spectra, improving product inspection and reducing false negatives in manufacturing.
Implementation Method 1
a spectroscopic image obtained by imaging a first type sample having an attached substance attached to a base with a spectroscopic camera
Data Source
AI summary
An attached substance determination method causes a computer to determine whether or not an attached substance is attached to an inspection target object, in which the computer includes at least one processor, and the at least one processor is configured to (a) acquire, as learning data, a spectroscopic image obtained by imaging a first type sample having the attached substance attached to a base with a spectroscopic camera and a spectroscopic image obtained by imaging a second type sample having no attached substance attached to the base with the spectroscopic camera, in which spectroscopic images of a plurality of kinds of the first type samples having different kinds of the bases and different kinds of the attached substances and spectroscopic images of a plurality of kinds of the second type samples having different kinds of the bases are acquired as the learning data, (b) generate, based on the learning data, a determination model with a spectroscopic image of the inspection target object as an input and a determination result relating to presence or absence of the attached substance as an output, (c) acquire the spectroscopic image of the inspection target object, and (d) input the spectroscopic image of the inspection target object to the determination model and determine the presence or absence of the attached substance based on the determination result output from the determination model.