The data and findings discussed here come from the cited paper and do not represent test results from VISPEK equipment.
Does a duck egg origin model perform as well on a conveyor as it does in offline testing? The study reports both a model comparison and a separate online equipment trial, making the gap visible. When interpreting recognition rates, the sample origins, operating conditions and timing scope all matter.
Summary
The research uses the transmission spectra of duck eggs from multiple sources to train a classification model, and then uses another set of duck eggs to test the online device. The online accuracy is lower than the highest score in the research test set, and the two cannot be mixed; the source categories are limited, and the detection time of a single piece does not represent the long-term stable production capacity of the entire line. Expansion to other production areas still needs to be re-verified.
(AI-generated conceptual illustration for context; it is not data from the paper, a photograph of the research equipment, or a VISPEK product diagram.)
Why It Matters
Reporting offline and online results separately gives egg-sorting developers a more realistic basis for comparing classification performance and processing speed.
Methods and Reference Measurements
Study and collect visible and near-infrared transmission data of duck eggs from three sources, and build a light shielding, roller conveying, light source and spectrum collection system. First compare smoothing and scattering correction, use SPA or CARS to filter wavelengths, and then build random forest, support vector machine and convolution classification models. The model comparison was completed with 182 eggs for training and 79 eggs for testing. An additional 30 duck eggs from three sources were used to run the online sorting device, and the classification performance and detection time of each egg were recorded. The results of the online device trials and the study test set are recorded separately.
Wavelengths and Acquisition Parameters
| Band/parameter | Original meaning and usage instructions |
|---|---|
| Instrument 200–1100 nm; this study analyzed 400–1100 nm | Collection/analysis scope of this study |
| SNV-SPA eight points: 439.45, 522.24, 557.01, 583.55, 610.9, 710.45, 722.79, 1075.42 nm | Original text features and evidence description |
Key Findings
In the 79-piece test, the CNN reached a maximum of 100%; the other 90 online devices tested the CNN accuracy rate of 94.44%, 0.5 s/piece, SVM91.11%, and RF90%. You cannot mix offline and online indicators.
Limitations and Interpretation
The conclusions from the three sources do not cover all production areas; the testing time of each model is not equal to the long-term stable production capacity of the entire line.
Image copyright belongs to the paper’s original authors. License: CC BY 4.0.
Source Paper
Li Q, Zhou W, Wang Q, Fu D. (2023). Research on Online Nondestructive Detection Technology of Duck Egg Origin Based on Visible/Near-Infrared Spectroscopy. Foods. https://doi.org/10.3390/foods12091900
Contact Us
Whether portable equipment can be used for similar detection needs to be evaluated based on samples and experimental conditions, and the specific effect still needs to be verified through actual testing. Welcome to contact VISPEK to exchange your sample and testing needs, jointly evaluate experimental plans, and explore the feasibility of using our company's equipment for actual measurement and verification.
- Email: contact@vispek.cn
- Website: https://www.vispek.cn
VISPEK products shown for illustration; these were not the devices used in the cited study.