The data and findings discussed here come from the cited paper and do not represent test results from VISPEK equipment.
Can point NIR measurements identify apple bruises with fewer wavelengths? The study compares classifiers on artificially bruised apples and reduces the spectral variables required. Its reported precision is distinct from overall accuracy, and induced bruises do not cover every kind of mechanical damage.
Summary
The study applied controlled bruises to several apple varieties, collected noncontact near-infrared reflectance spectra, and compared wavelength-selection methods and classifiers. A model using relatively few wavelengths achieved test precision similar to a full-spectrum model, but precision is not overall accuracy. The induced bruises do not represent all mechanical damage, and the selected wavelengths also included information beyond the target shortwave range.
(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
The work offers a sample-preserving route for developing spectroscopic spot checks of apple bruises.
Methods and Reference Measurements
A total of 214 apples from Golden Delicious, Granny Smith, and Royal Gala were purchased, and steel balls were dropped from three heights, causing bruises. The Matrix-F non-contact transmitter collects 800-2500 nm diffuse reflectance spectra on both sides of the fruit body, averaging 64 scans each time. After standardization, RFE/SFS is used to select waves and test multiple classifiers.
Wavelengths and Acquisition Parameters
| Band/parameter | Original meaning and usage instructions |
|---|---|
| 800–2500 nm | Original study modeling scope; see Methods and Results for complete wavelength selection. |
Key Findings
The linear classification method and SVM select no more than 40 wavelengths, and the test precision is 0.79–0.86, which is equivalent to the error range of the full-spectrum model. Precision is the accuracy rate, not the overall accuracy rate.
Limitations and Interpretation
The reported precision is the accuracy rate rather than the overall accuracy rate; manual bumps do not cover all mechanical damage, and the wave selection results include about 1900 nm.
Band and model boundary: Band overlap: original study 800–2500 nm, including information above 1700 nm; only used for application inspiration, and has not proven that the target range alone maintains performance.
Image copyright belongs to the paper’s original authors. License: CC BY 4.0.
Source Paper
Nturambirwe JFI, Hussein EA, Vaccari M, Thron C, Perold WJ, Opara UL. (2023). Feature Reduction for the Classification of Bruise Damage to Apple Fruit Using a Contactless FT-NIR Spectroscopy with Machine Learning. Foods (Basel, Switzerland). https://doi.org/10.3390/foods12010210
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.