Food Product Sorting via Spectral Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging methods for food products, including cameras and machine learning algorithms, struggle to accurately determine the quality of food products being conveyed in a factory setting, as simple images may not provide sufficient information.
Innovation Solution
A device and method utilizing a line-scan dispersive spectrometer to acquire spectra of food products, combined with machine learning algorithms trained to classify these spectra into categories indicative of food parameters, to sort and label food products accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If simple camera images are used to determine food product quality, then the device complexity is low, but the measurement precision is insufficient
Solution Approach 1:
The patent replaces simple mechanical camera imaging with a spectrometer-based optical sensing system that captures spectral data. This substitution enables precise chemical and physical property measurement of food products through light interaction, achieving high measurement precision while maintaining automated operation. The spectrometer system provides detailed spectral signatures that reveal food quality parameters beyond what simple images can detect.
Solution Approach 2:
The patent transforms the measurement approach by changing from capturing spatial image data to capturing spectral parameter data. The system measures multiple wavelength parameters simultaneously, converting food product properties into spectral signatures. This parameter transformation enables precise quality determination through machine learning algorithms that analyze spectral patterns, achieving high accuracy without proportionally increasing device complexity.
2Measurement precision
If spectral data acquisition is implemented for each segment, then the measurement precision improves, but the productivity decreases due to detailed analysis requirements
Solution Approach 1:
The patent divides the food product into multiple segments along its length, with each segment independently analyzed by the spectrometer. The system captures spectral data for each segment and applies machine learning classification to determine quality characteristics. This segmentation approach enables precise localized quality assessment while maintaining high throughput through automated parallel processing of multiple segments.
Solution Approach 2:
The system performs preliminary spectral data acquisition and machine learning classification for each segment before the physical sorting action. By pre-classifying segments based on their spectral signatures, the system prepares sorting decisions in advance, enabling high-speed automated sorting without sacrificing measurement precision. The classification results guide subsequent sorting operations efficiently.
3Reliability
If machine learning algorithms are trained to classify spectra, then the reliability of food parameter classification improves, but the loss of time increases during training
Solution Approach 1:
The patent implements preliminary training of machine learning algorithms using historical spectral data and known food quality outcomes. The system pre-trains classification models offline before deployment, establishing reliable spectral-to-quality mappings in advance. This preliminary action ensures high classification reliability during operation while minimizing real-time processing delays, as the trained models can quickly classify new spectral data without requiring extensive training during production.
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
The system effectively classifies food products on a segment-by-segment basis or in groups, enabling precise sorting and labeling based on food parameters, thereby improving the accuracy of food quality determination.
Implementation Method 1
at least one line-scan dispersive spectrometer configured to acquire respective spectra of the food products
Implementation Method 2
acquire respective spectra of the food products for (e.g. each of) a plurality of segments of a line
Data Source
AI summary
Devices, systems and methods for sorting and labelling food products are provided. Respective spectra of food products for a plurality of segments of a line are received at a controller from at least one line-scan dispersive spectrometer configured to acquire respective spectra of the food products for the plurality of segments of the line. The controller applies one or more machine learning algorithms to the respective spectra to classify the plurality of segments according to at least one of one or more food parameters. The controller controls one or more of a sorting device and a labelling device according to classifying the plurality of segments to cause the food products to be one or more of sorted and labelled according to the at least one of the one or more food parameters.