Estimation of the sensory properties of black tea samples using non-destructive near-infrared spectroscopy sensors
FOOD CONTROL, cilt.142, ss.109260, 2022 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 142
- Basım Tarihi: 2022
- Doi Numarası: 10.1016/j.foodcont.2022.109260
- Dergi Adı: FOOD CONTROL
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, BIOSIS, CAB Abstracts, Food Science & Technology Abstracts, Index Islamicus, Veterinary Science Database
- Sayfa Sayıları: ss.109260
- Anahtar Kelimeler: Cupping test, NIR chemometric Models, Python, Non-destructive sensors, PCR, PLSR, TOTAL POLYPHENOLS CONTENT, GREEN TEA, PATTERN-RECOGNITION, QUALITY, DISCRIMINATION, NIR, VARIETIES, SPECTRA, PCA, CLASSIFICATION
- Süleyman Demirel Üniversitesi Adresli: Hayır
Özet
The quality characteristics of black tea are routinely assessed before it is purchased, blended and marketed to ensure its quality and value. Although some of these quality characteristics can be measured analytically, others need to be determined as sensory scores following cupping tests conducted by tea experts. However, most of these analyses (especially the sensory ones) require high training and expertise, are time-consuming and prone to human error. Therefore, in this study, non-destructive spectral sensors were combined with chemometric methods to rapidly measure the results of the cupping test (appearance, body, colour and overall quality) and some other important sensory quality attributes (bulk density, cellulose, water extract and moisture) of black tea samples. A total of 54 black tea samples from Turkiye were analysed in three different NIRS (near-infrared spectroscopy) devices (MicroNIRTM 1700, Matrix-F FT-NIR and NIRS DS 2500). Partial Least Squares Regression (PLSR) and Principal Components Regression (PCR) with stepwise variable elimination were used as regression algorithms to develop the NIR calibrations. As a result, PLSR provided slightly superior estimates with R2cv be-tween 0.83 and 0.97 and RPDcv between 2.47 and 5.79 for sensory traits. For analytical traits, model statistics for PLSR ranged between 0.66-0.89 and 1.72-3.08 for R(cv )(2)and RPDcv, respectively. These results suggest that PLSR combined with FT-NIR technology may be promising for rapid and economical evaluation of sensory (cupping test) scores and related properties for its use in the tea industry.