Genetic-Algorithm Spectral Wavelength Selection for Product Discrimination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing spectroscopic methods for discriminating between normal and abnormal products face challenges such as increased information acquisition and measurement time when the number of wavelengths is increased, without necessarily improving discrimination accuracy.
Innovation Solution
A determination method using a genetic algorithm to select optimal wavelengths and irradiation conditions for spectroscopic spectra, employing models like Fisher linear discrimination and Mahalanobis distance to improve discrimination accuracy while reducing the number of wavelengths used.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of wavelengths is increased to improve discrimination accuracy, then discrimination accuracy is improved, but measurement time and information acquisition time increase
Solution Approach 1:
The patent extracts only the essential wavelengths from the full spectrum that are necessary for discrimination. By identifying and removing redundant wavelengths, the system maintains discrimination accuracy while significantly reducing the number of wavelengths that need to be measured, thereby reducing measurement time.
Solution Approach 2:
The patent segments the full spectral range into multiple wavelength bands and selectively measures only specific wavelengths within these bands that contribute most to discrimination accuracy. This segmentation approach allows the system to avoid measuring all wavelengths while still capturing the essential information needed for accurate discrimination.
2Measurement precision
If the number of wavelengths is increased to improve discrimination accuracy, then discrimination accuracy is improved, but the amount of information to be acquired increases
Solution Approach 1:
The patent extracts only the critical wavelength information necessary for discrimination from the full spectrum. By identifying and eliminating redundant wavelength data, the system reduces the total amount of information that needs to be acquired and processed while preserving the essential discriminative features.
Solution Approach 2:
The patent performs preliminary analysis to identify which wavelengths are most informative for discrimination before actual measurement. This preliminary action allows the system to pre-determine the optimal set of wavelengths to measure, avoiding the acquisition of unnecessary information in the first place.
3Measurement precision
If the number of wavelengths is increased to improve discrimination accuracy, then discrimination accuracy may be improved, but the system complexity increases
Solution Approach 1:
The patent extracts only the essential wavelengths needed for discrimination, removing unnecessary wavelengths from the measurement system. This extraction simplifies the optical system by reducing the number of wavelength channels, detectors, and associated processing requirements while maintaining discrimination accuracy.
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
This approach enhances discrimination accuracy by selecting wavelengths and conditions that maximize separation between normal and abnormal products, reducing the number of wavelengths required and measurement time.
Implementation Method 1
irradiating an inspection target object with near-infrared light and determining whether or not the inspection target object is a normal product or an abnormal product by using a spectroscopic spectrum of reflected light from the inspection target object
Data Source
AI summary
Provided is a determination method that includes obtaining measurement data, selecting one or more second wavelengths from a plurality of first wavelengths including at least one of a plurality of measurement wavelengths to generate a plurality of individuals, by using a genetic algorithm, inputting, to a first model learned to reproduce a correct answer label of a target object, the measurement data of the target object belonging to a remaining group and a second spectroscopic spectrum determined by the second wavelength to discriminate a label of the target object belonging to the remaining group, for each of the plurality of individuals, and determining whether or not to use the second wavelength as the wavelength of the spectroscopic spectrum for discrimination based on a rate at which the label is correctly discriminated.