Non-destructive textural quality assessment of peaches and nectarines using near-infrared spectroscopy integration time
The article addresses the challenge of non-destructively assessing the ripeness and quality of peaches and nectarines, as conventional methods often require invasive testing that can compromise the product. The main objective is to develop a non-destructive method based on near-infrared (NIR) spectr...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-12-01
|
| Series: | Applied Food Research |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2772502225005074 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The article addresses the challenge of non-destructively assessing the ripeness and quality of peaches and nectarines, as conventional methods often require invasive testing that can compromise the product. The main objective is to develop a non-destructive method based on near-infrared (NIR) spectroscopy integration time to classify fruit by textural properties throughout the postharvest process. The working hypothesis suggests that NIR integration time, influenced by light scattering properties, is related to the textural characteristics of the fruit.For the study, 725 samples of ten varieties of peaches and nectarines were examined at three stages: just harvested, during the postharvest protocol, and ready-to-eat. Samples underwent both non-destructive NIR spectroscopy (with integration times adjusted for optimal signal-to-noise ratio) and destructive tests, including the Magness-Taylor firmness test and juiciness measurements on absorbent paper. A principal component analysis (PCA) was used to analyze textural variability, and discriminant function analysis (DFA) and artificial neural networks (ANNs) were applied to classify fruit firmness levels.PCA effectively defined a new variable, PC1, that captured the variance in fruit texture, correlating positively with juiciness and negatively with firmness. DFA achieved classification accuracies of 78.22 % for peaches and 72.33 % for nectarines, while ANNs further improved accuracy, achieving over 90 % for both fruits. These findings indicate that integration time, combined with machine learning, is a feasible, efficient method for non-destructive ripeness classification. In conclusion, the integration time parameter shows strong potential as a rapid, non-invasive tool for assessing fruit quality, offering a practical, cost-effective solution for postharvest applications. |
|---|---|
| ISSN: | 2772-5022 |