No abstract

a Tea quality used to be traditionally evaluated by sensory panels. Their expertise is considerably better than that of the average consumers. However, there can be definite subjectivity in the judgment of sensory panels. A new direction is advanced sensory study that integrates sensory assessment with instrumental evaluation. The aim of this paper is to distinguish five Sri Lanka black teas from plantations of different geographical origins. From the tea samples, infusions were prepared with standard procedure. Three different test methods were applied: electronic nose, electronic tongue and sensory assessment. In the case of the electronic nose only the volatile components were evaluated. The Alpha Astree II potentiometric electronic tongue analyzed the liquid phase, while in the sensory profiling all perceptible attributes were tested. Data were analyzed with principle component analysis (PCA), linear discriminant analysis (LDA) and in the case of sensory analysis with one-way ANOVA. The partial least square regression (PLS) was used to estimate the sensory attributes. As a final result, we found that LDA maps show the quality differences of teas grown at different altitudes and geographical location of growing regions.

No abstract

Near Infrared Hyperspectral Imaging (NIRHSI) is an emerging technology platform that integrates conventional imaging and spectroscopy to attain both spatial and spectral information from an object. Two important problems in NIRHSI are those of data load and unserviceable pixels in the NIR sensor. Hyperspectral imaging experiments generate large amounts of data (typically > 50 MB per image), which tend to overwhelm the memory capacity of conventional computer systems. This inhibits the utilisation of NIRHSI for routine online industrial application. In general, approximately 1% of pixels in NIR detectors are unserviceable or ‘dead’, containing no useful information. While this percentage of pixels is insignificant for single wavelength imaging, the problem is amplified in NIRHSI, where > 100 wavelength images are typically acquired. This paper describes an approach for reducing the data load of hyperspectral experiments by using sample-specific vector-to-scalar operators for real time feature extraction and a systematic procedure for compensating for ‘dead’ pixels in the NIR sensor. The feasibility of this approach was tested for prediction of moisture content in carrot tissue.

In this article a trained sensory panel evaluated 6 flavored mineral water samples. The samples consisted of 3 different brands, each with 2 flavors (pear-lemon grass and josta berry). The applied sensory method was profile analysis. Our aim was to analyze the sensory profiles and to investigate the similarities between the sensitivity of the trained human panel and an electronic tongue device. Another objective was to demonstrate the possibilities for the prediction of sensory attributes from electronic tongue measurements using a multivariate statistical method (Partial Least Squares regression [PLS]). The results showed that the products manufactured under different brand name but with the same aromas had very similar sensory profiles. The panel performance evaluation showed that it is appropriate (discrimination ability, repeatability, and panel consensus) to compare the panel's results with the results of the electronic tongue. The samples can be discriminated by the electronic tongue and an accurate classification model can be built. Principal Component Analysis BiPlot diagrams showed that Brand A and B were similar because the manufacturers use the same aroma brands for their products. It can be concluded that Brand C was quite different compared to the other samples independently of the aroma content. Based on the electronic tongue results good prediction models can be obtained with high correlation coefficient (r(2) > 0.81) and low prediction error (RMSEP < 13.71 on the scale of the sensory evaluation from 0 to 100).

No abstract

No abstract

Reliability and validity of sensory data is an important issue in scientific researches. If sensory analysis is performed in an analytical approach, the resulting data will show a similar structure to the chemical analyses. In the present paper the authors have used a complex approach to evaluate the performance of a sensory panel. The tested samples were black tea batches from different plantations of Sri Lanka. Profile analysis was applied to identify the odor profiles of the samples. Sensory profile data was submitted to two novel techniques of panel performance evaluation. GCAP (Gravity Center Area/Perimeter) is based on the profile polygons of the individual assessors. If the area/perimeter ratio of two panelists' profiles is similar and the gravity center is located near to each other, the panelists performed the tests consistently. CRRN (Compare Ranks with Random Numbers) is applicable not only to sensory data but also to other field of chemometrics. The essence of CRRN method is based on the evaluation of an 'average' vector, corresponding to the coordinate-averages of the measured points, and on a produced random vector series of the same dimension as the measured points. Sensory and e-nose data were evaluated with principal component analysis, cluster analysis and linear discriminant analysis. Partial least square regression and support vector machine regression were used to predict sensory data with electronic nose results. Prediction by support vector machine gave close correlation between the results of electronic nose measurement and odor attributes.