Articles
Detection of fruit juice adulteration by laser backscattering imaging
Article number
1382_9
Pages
69 – 76
Language
English
Abstract
Water dilution is the most common practice of juice adulteration, raising customers concern about misleading quality and resulting in financial damage.
This study focused on investigating the applicability of the non-destructive laser light backscattering (LLB) imaging technique on assessment of adulteration of fruit juices with different levels of dilution with water.
Orange, orange with bits, pink grapefruit and cranberry juices of 100% fruit juice were used in the experiments and samples were diluted to 95, 90, 85 and 80%. Six low power laser modules, emitting at 532, 635, 780, 808, 850, 1064 nm wavelength, were used.
The LLBI profile was extracted from 12 bit raw images and described with the peak widths at three selected intensity levels (D75%, D50%, D25%) and their ratios; the illuminated area at selected intensity level of 50% (A50%), the ring of 25-75% and their ratio.
The ring of 25-75% obtained the highest sensitivity to juice dilution at all wavelengths based on ANOVA F-value (p<0.001). Diluted juices were successfully distinguished by linear discriminant analysis (LDA) from 5% (v/v) adulteration.
Moreover, the type of juices at the same dilution level were recognized by LDA with the correct classification rate above 91.6%. Partial least squares (PLS) regression was successfully applied to predict the Brix value of adulterated juices with the performance of determination coefficient (R2CV) of 0.75 and estimation error (RMSECV) of 0.542%. According to the results, laser backscattering imaging is a promising non-invasive technique for quality control of fruit juices.
This study focused on investigating the applicability of the non-destructive laser light backscattering (LLB) imaging technique on assessment of adulteration of fruit juices with different levels of dilution with water.
Orange, orange with bits, pink grapefruit and cranberry juices of 100% fruit juice were used in the experiments and samples were diluted to 95, 90, 85 and 80%. Six low power laser modules, emitting at 532, 635, 780, 808, 850, 1064 nm wavelength, were used.
The LLBI profile was extracted from 12 bit raw images and described with the peak widths at three selected intensity levels (D75%, D50%, D25%) and their ratios; the illuminated area at selected intensity level of 50% (A50%), the ring of 25-75% and their ratio.
The ring of 25-75% obtained the highest sensitivity to juice dilution at all wavelengths based on ANOVA F-value (p<0.001). Diluted juices were successfully distinguished by linear discriminant analysis (LDA) from 5% (v/v) adulteration.
Moreover, the type of juices at the same dilution level were recognized by LDA with the correct classification rate above 91.6%. Partial least squares (PLS) regression was successfully applied to predict the Brix value of adulterated juices with the performance of determination coefficient (R2CV) of 0.75 and estimation error (RMSECV) of 0.542%. According to the results, laser backscattering imaging is a promising non-invasive technique for quality control of fruit juices.
Authors
H.X. Mac, N.T.T. Ha, T.T. Pham, L.P.L. Nguyen, L. Baranyai
Keywords
diffuse reflectance imaging, food quality control, laser light, non-destructive
Online Articles (30)