Near-infrared spectroscopy with linear discriminant analysis for green ‘Robusta’ coffee bean sorting
The present work investigated the feasibility of near-infrared (NIR) spectroscopy for separation of good quality green ‘Robusta’ coffee beans from defective (broken beans, beans with parchment, and beans with husk) and contaminated beans (faecal matter and soil) by single bean measurement. Linear di...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Journal Contribution |
| Language: | English |
| Published: |
2023
|
| Subjects: | |
| Online Access: | http://agris.upm.edu.my:8080/dspace/handle/0/23300 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The present work investigated the feasibility of near-infrared (NIR) spectroscopy for separation of good quality green ‘Robusta’ coffee beans from defective (broken beans, beans with parchment, and beans with husk) and contaminated beans (faecal matter and soil) by single bean measurement. Linear discriminant analysis using principal components from principal component analysis (PCA-LDA) as variables was used as a supervised method for the classification. It was found that smoothing pre-treatment applied to the spectra was suitable for the classification, with the highest classification accuracy of 97.5%. The present work indicated that NIR spectroscopy coupled with appropriate chemometric methods could be an efficient tool for coffee bean sorting. |
|---|