1 - Colour Training and Colour Differences Thresholds in Orange Juice

Publish in



Please download to get full document.

View again

of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Food Quality and Preference 30 (2013) 320–327 Contents lists available at SciVerse ScienceDirect Food Quality and Preference journal homepage: www.elsevier.com/locate/foodqual Colour training and colour differences thresholds in orange juice Rocío Fernández-Vázquez a, Carla M. Stinco a, Dolores Hernanz b, Francisco J. Heredia a, Isabel M. Vicario a,⇑ a b Food Colour & Quality Laboratory, Department of Nutrition & Food Science, Universidad de Sevilla, Facultad de Farmacia, 41012 Sevilla, Spai
  Colour training and colour differences thresholds in orange juice Rocío Fernández-Vázquez a , Carla M. Stinco a , Dolores Hernanz b , Francisco J. Heredia a , Isabel M. Vicario a, ⇑ a Food Colour & Quality Laboratory, Department of Nutrition & Food Science, Universidad de Sevilla, Facultad de Farmacia, 41012 Sevilla, Spain 1 b Department Analytical Chemistry, Universidad de Sevilla, Facultad de Farmacia, 41012 Sevilla, Spain a r t i c l e i n f o  Article history: Received 1 October 2012Received in revised form 3 April 2013Accepted 31 May 2013Available online 12 July 2013 Keywords: ColourOrange juiceSensory trainingColour differences a b s t r a c t This study was aimed at training a panel of assessors to evaluate specifically orange juice colour, and toestablishthecolourdifferencethresholdinorangejuiceforatrainedanduntrainedpanel.Panellistswerefirst preselected using Farnsworth–Munsell 100-Hue Test and then trained with a specific method fororange juice colour. This training allowed assessors to evaluate visually orange juice samples in hueand intensity. The final selection of assessors was a panel of 8 trained observers with reproducibilityand repeatability, and a significant discrimination among the samples (  p  <0.05) . On the other hand, commercial orange juices were evaluated both instrumentally by image analysisandvisuallybythetrainedpanel, andtheuntrainedpanel. Instrumental colour measurements andvisualevaluation were correlated. Values around 1.5 and 2.8 CIELAB units could be consider the threshold forcolour differences between two orange juices for the trained and untrained panel, respectively.   2013 Elsevier Ltd. All rights reserved. 1. Introduction Colourisoneofthemostimportantvisualattributesinfoodandusuallyis thefirst oneevaluatedbyconsumersandisassociatedtothe concept of quality (Huggart, Petrus, & Buzz Lig, 1977;Pangborn, 1960; Tepper, 1993). In orange juices, the natural brightcolouris consideredoneof theirmainadvantagesoverother juices(Barron, Maraulja, & Huggart, 1967) and has attached greatimportance since some studies have probed that it may influenceflavour perception and other quality attributes (Fernández-Vázquez et al., 2012; Tepper, 1993).Colour can be evaluated by instrumental or visual analysis.Humans and instruments measure colour in different ways.Human perception of colour is based on responses of photorecep-tors inthe retinaof the eye andthe way theyareinterpretedwith-in the brain. These perceived colours are often characterised byphysical scientists using three dimensions: lightness, hue andchroma.Instruments, ontheotherhand, arecapableof seeingpurevalues of the colorimetric coordinates CIELAB  L ⁄ ,  a ⁄ , and  b ⁄ . Nowa-days, there are new advances in image acquisition technology thatoffer the possibility of using technically sophisticated apparatusavailable at relatively low cost to evaluate colour in terms of mil-lions of pixels. In comparison with the traditional light sensors,themainadvantageisthattheyallowmakingadetailedevaluationof a wider area of any food product, with inhomogeneous colourpossible. Every different colour in the image of the analysed foodmatrix can be accounted for by one or more pixels (Antonelliet al., 2004). Furthermore, it is based upon digital cameras, whichcan quickly capture images in digital format (DigiEye  ) (Luo, Cui,& Li, 2001) and offers a more reliable measurement of the foodcolour, which can be correlated with sensory analysis and othercolour measurements (Fernández-Vázquez, Stinco, Melendez-Martínez, Heredia, & Vicario, 2011).Anyway,colourmeasurementusuallyrequiresinstrumentsthatare not always available in small and medium size companies andvisual evaluation could be an alternative. Human colour vision is aquite complex process and colour is undoubtedly a perception, avirtual property of the material. In order to use the visual analysisas an objective quality control, it is necessary to standardize themeasurement conditions to be able to compare with the instru-mental measurement. Previous studies have shown that a goodcorrelation can be achieved when the instrumental and sensorymeasurements are done considering different aspect such as back-ground, surround or illumination (Fernández-Vázquez et al., 2011;Meléndez-Martínez, Vicario, & Heredia, 2005).Although colour evaluation is included in many sensory studies(Calvo, Salvador, & Fiszman, 2001; Frata, Valim, & Monteiro, 2006;Poelman & Delahunty, 2011), there are very few studies whichspecially train the panellists to do the visual evaluation of foodwith more details. An example of an specific training in visualevaluation was done by Gambaro, Giménez, and Burgueño (2001)for strawberry yoghourt. Based on this experience, we haveparticularly trained a panel to evaluate orange juice colour in areproducible and repeatable way.On the other hand, the evaluation of colour differences has hada high interest for long time. Specifically, ‘just noticeable 0950-3293/$ - see front matter   2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.foodqual.2013.05.018 ⇑ Corresponding author. Tel.: +34 954556339; fax: +34 954557017. E-mail address:  vicario@us.es (I.M. Vicario). 1 http://www.color.us.es. Food Quality and Preference 30 (2013) 320–327 Contents lists available at SciVerse ScienceDirect Food Quality and Preference journal homepage: www.elsevier.com/locate/foodqual  differences’ have been very important in the development of thecolorimetry. There are equations to find out the colour differencebetween two stimuli in the CIELAB space, which are mathematicalexpressions which allows us to obtain the number  D E  ab . This is thepositive number which stays invariable when the products areexchanged (Melgosa, Pérez, Yebra, Huertas, & Hita, 2001). At thepresent time, calculation of colour differences has many applica-tions in colorimetry, such as the reproducibility of colour inmanufactured products and communication systems, or in thestudy of the colour fading in food, works of art, etc. or more re-cently to determine colour tolerance in orange juice (Wei, Ou,Luo, & Hutchings, 2012).One of the key problems in the visual evaluation is establishingthethresholdforcolourdifferences.Previousstudieshaveexploredcolourthresholdusingcolourstandards(Berns, Alman, Reniff,Sny-der, &Balonon-Rosen, 1991). An attempt to stablish the thresholdsfor visual discrimation between wines was also published byMartínez, Melgosa, Pérez, Hita, and Negueruela (2001). RecentlyWei et al. (2012) established the colour of an ideal orange juiceand the colour tolerance, using a digital display. However, so far,literature on orange juice colour does not provide data on the col-our differences that can be visually detected between two orange juices (based on real samples) by consumers, although this typeof information could be very useful for the orange juice industry.The objectives of this study were: (1) to train a panel of asses-sors to evaluate specifically the orange juice colour, and (2) tostudythevisuallyperceivedcolourdifferencebytheobservers’pa-nel (trained and untrained) in a complete range of orange juices of differentcolourstoestablishthecolourdifferencethresholdinthispopular beverage. 2. Material and methods  2.1. Instrumental colour measurement  The DigiEye imaging system was used to capture the digitalimages (Luo et al., 2001). The latter system includes a calibrateddigital camera 10.2-megapixel Nikon D80 (Nikon Corporation,Tokyo, Japan) and an objective Nikkor 35-mm f/2D (NikonCorporation), a colour sensor for display calibration, and an illumi-nation box designed by VeriVide Ltd. (Leicester, UK) (Fig. 1). Forobjective colour specifications, the samples were placed in 75mL capacity transparent plastic bottles (Fig. 2) and measured againsta grey surround ( L ⁄ =50) and white background. Digital imageswere made in order to obtain the total appearance of juice atdepths observed by consumers. In these measurements, the sam-ples were illuminated by a diffuse D65 simulator. For obtainingCIELAB coordinates, we used the DigiFood software (Heredia,González-Miret,Álvarez,&Ramírez,2006),whichallowsthetrans-formation of RGB values into the CIELAB colour parameters, basedoncomputationalsolutions(León, Domingo, Franco, &León, 2006).From the CIELAB uniform colour space, the psychophysicalparameters chroma ( C   ab ) and hue ( h ab ) are defined as: C   ab  ¼  ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  a  ð Þ 2 þ  b  ð Þ 2 q   ; h ab  ¼  arctan  b  = a  ð Þ Chroma ( C   ab ) is used to determine the degree of difference of a huein comparison with a grey colour with the same lightness, and isconsidered the quantitative attribute of colourfulness. Hue ( h ab ) isthe attribute according to which colours are usually defined as red-dish, greenish, etc. and is used to define the difference of a colourwith reference to a grey colour with the same lightness. This attri-bute is related to the differences in reflectance at different wave-lengths and is considered the qualitative attribute of colour.Colourdifferences ( D E  ab ) were calculated as the Euclidean distancebetween two points in the 3-D space defined by  L ⁄ ,  a ⁄ and  b ⁄ : D E   ab  ¼  ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  D L  ð Þ 2 þ  D a  ð Þ 2 þ  D b  ð Þ 2 q   2.2. Colour training  Briefly, aprotocol wasdesignedforthe selectionandtrainingof assessors based on the methodology proposed by Gambaro et al.(2001). First, a preliminary panellist selection was made usingthe Farnsworth-Munsell 100-Hue Test. Then, those panellistswho did not present a good skill to discriminate light differencesin tone and intensity using blended of colouring dilutions were 1 4 32 251 4   32 25 1: Domed cabinet2: Fluorescent tubes. D65 simulator 3: Sample 4: Digital camera 5: PC Digifood® Software Fig. 1.  Scheme of the DigiEye System. R. Fernández-Vázquez et al./Food Quality and Preference 30 (2013) 320–327   321  rejected. Finally, the selected panel was training using two com-mercial samples and a serie of dilutions of one of them.  2.2.1. Panel selection A panel of 12 assessors were recruited from students and staff at the University of Seville and then preselected according to theirnormal colour vision following ISO 11037 (1999). The Farnsworth-Munsell 100-Hue Test for the examination of Colour Discrimina-tion (Farnsworth 1957) was used to verify the normal vision. It al-lows to separate persons with normal colour vision into classes of superior, average and low colour discrimination, and to measurethe zones of colour confusion in colour defective persons.To determine the ability to discriminate among slight tone andintensity differences in orange colour, aqueous orange-colouredsolutions were prepared using two food dyes (red and yellow fooddyes fromMcCormick, Spain S.A) blended in different proportions:500 and 12,000 l L/L of red and yellow food dye respectively, foryellowish dilutions; 1,250 and 12,000 l L/L of red and yellow fooddye for orangish dilutions; and 3,000 and 12,000 l L/L of red andyellow food dye for reddish dilutions. Ten aqueous solutions of each blend of colourings were prepared (100, 78, 47, 36, 29, 22,17, 14, 10, and 5%). These solutions (75mL) were placed in bottlesof transparent plastic and coded with 2 digit random numbers.The test was carried out using a VeriVide CAC Portable cabinet(dimension of viewing area: 635mm width, 280mm height, and280mmdepth)tocontrolilluminationandobservationconditions.D65 was used as source of illumination (the same used in theinstrumental measurements) (CIE, 2007), a white backgroundand a grey surround were selected to simulate the objective mea-surements made by image analyses (Stinco et al., 2012).The evaluation sessions were organised as follows: in a firststage (hue classification) assessors dealt with the samples ( n  =3)correspondingtotheeachdilutionlevel separatelyandwereaskedto sort them in yellowish, orangish, and reddish hues (10 evalua-tions). In a second stage (intensity ranking), the assessors weregiven the whole dilution series (yellowish, orangish and reddish)separately, and they were asked to rank them according to theincreasing intensity.To select the assessors, the criteria used were: (1) reject thosewho were unable to accomplish sorting the tubes into the threetonegroupsand(2)rejectthosewhoseSpearman’s rankedcorrela-tion coefficients (  p ) of sensory ranking versus colour concentra-tions were not significant (  p  >0.05) (O’Mahony, 1986).  2.2.2. Panel training  Two nonstructured 10cm long scales, anchored at the end,were used to train the assessors in colour evaluation. The colourattributes trained were hue and intensity. In this study, chromaand lightness were not considered separately as individual attri-butes because in previous studies it was observed that panellistshad difficulties to understand and evaluate chroma (Fernández-Vázquez et al., 2011). For this reason, intensity was assayed asthe best way to evaluate both parameters visually. Hue was evalu-ated from yellowish to reddish and intensity was evaluated fromlow to high.A collection of samples were selected to encompass the fullrange of colour intensity and hue in commercial orange juice.Two commercials orange juice samples (COJ I and COJ II), and sixsamples prepared from dilutions of COJ II (6%, 10%, 30%, 50%,60%, and 80%) were used. These samples were evaluated in15min sessions, and at the end of each session a meeting of 30min was done by the leader of the panel and all the panelliststo unify the criteria of evaluation. In each session, assessors evalu-ated duplicates of the commercial samples and a couple of dilutedsamples.The samples evaluated in each session were the same for allassessors,beingtheorderrandomisedacrossassessors.Thisdesignallowedtrainingoftheassessorsinevaluatingorangecolourinten-sity and hue as well as determining their reproducibility and per-formance consistency. A three factor ANOVA (assessor, sample,and repetition) for samples COJ I and COJ II, and a two factor ANO-VA (assessor and sample) for these two samples and all the dilu-tions were performed on the data obtained (O’Mahony 1986).According to the results obtained, those assessors with the highestvariance and greater judgement dispersion were withdrawn fromthe panel.  2.3. Colour differences thresholds 2.3.1. Orange juices samples 16 commercial orange juices (5 from concentrated, 6 fromsqueezed oranges and stored at 4  C, and 5 from squeezed orangesand stored at room temperature) were purchased from differentsupermarkets in Spain. These samples were chosen in order tocollect the variety of the orange juices colour available in thesupermarket. Each sample was placed in 75mL capacity transpar-ent plastic bottles to measure its colour by image analyses andthen to evaluate the colour differences.  2.3.2. Sensory evaluation The samples were compared by pairs (120 comparisons) by thetrained panel of 8 observers with normal colour vision and previ-ously trained in colour discrimination experiments.Afterwards, eight panellists recruited also from students andstaff at the University of Seville, with normal vision (according totheFarnsworth-Munsell 100-HueTest) but nopreviousknowledgein colour science, repeated the experiment with the aim of establishing the colour difference threshold for untrainedobservers.120 pairs of samples were displayed on the centre of theVeriVide CAC 120 cabinet. Observers were situated 50cm in frontof the samples, with white background and grey surround.The task of each observer was to judge whether they couldnotice the colour difference between the two orange juice samples(1) or if they could not (0). They did the test twice per pair of samples in two different sessions: one with the couple of samples 74 mm48 mm29 mm74 mm48 mm29 mm Fig. 2.  Characteristics of the bottles used for containing the samples in theinstrumental colour measurements and the visual evaluation. (For interpretation of the references to colour in this figure legend, the reader is referred to the webversion of this article.)322  R. Fernández-Vázquez et al./Food Quality and Preference 30 (2013) 320–327   sidebyside(experimenta);andanotheronewiththesamplessep-arated by 15cm (experiment b). In this way, each observer in thetrainedpanel(8panellists)evaluatedthe120pairsofsamplesonce(960 judgments in total), similarly did the group of 8 untrainedobservers (1200 judgments in total). The final results were ex-pressed as ‘‘visual colour difference’’ ( D V  ) which was calculatedas the percentage of positive panellists’ responses (1).The visual judgments were made immediately after the instru-mental colourmeasurementsinorder to avoidthe colour variationof the samples.The correlation between the visually perceived and instrumen-tally measured colour differences were explored following proce-dures previously described elsewhere (Davidson & Friede, 1953;Kuehni,1976;Martínezetal.,2001;Strocka,Brockes,&Paffhausen,1983). The percentage of the positive colour differences perceivedby both panels ( D V  ) were plotted against the CIELAB colour differ-ences (( D E  ab ) instrumentally measured. Then, an S-shaped curve(  y  =  A /[1+exp( B  + Cx )]) was fitted using an iterative algorithm of successiveapproximationstothefunctionanditsderivatives, untilmaximising the value of   r  2 (Martínez et al., 2001). The softwareMATLAB R2011b (The MathWorks Inc., Natik, Massachusetts) wasused for this purpose. For the final difference threshold, the 50%of positive responses by the observers was consider as the typicalmeasurement of tolerance or acceptability of colour differencesperceived (Alman, Berns, Snyder, & Larsen, 1989; Berns et al.,1991; Qiao, Berns, Reniff, & Montag, 1998).  2.4. Data analysis All statistical analyses were performed using the the programStatistica 8 for Windows (StatSoft, 2007). 3. Results and discussion  3.1. Colour training  3.1.1. Panel selection The Farnsworth-Munsell 100-Hue Test was applied to theassessors and results showed that all passed the test withpunctuation lower than 48. This mean that some of them was inthe group of superior discrimination (scores lower than 16) andothers were in the group of average discrimination (scores lowerthan100).AnexampleoftheresultsisshowninFig.3.Theseverityofthedefectcanbegaugedbytheextentofthe‘bulge’,aseverede-gree of defect would show clear bipolarity with high error scores;moderate cases would show small ‘bulges’ and lower total errorscores; mild cases with good colour discrimination would showno ‘bulge’ and cannot be identified by this test.The food dye solutions used for tone separation and intensityranking were analysed by image analysis. The yellowish seriehad a lightness ranging from 56.02 to 76.17; the orangish from48.14 to 74.01 and finally the reddish serie from 43.51 to 70.11.Chroma ranges were 44.45–65.23, 46.57–62.29, and 44.12–60.48for the yellowish, the orangish and the reddish series, respectively.Finally, hue angle data ranged from58.1  to 95.9  inyellowish ser-ie, from37.0  to 91.3  in the orangish serie and from30.8  to 83.3  in the reddish serie (Fig. 4).After the panellists sorted the samples out and ranked themaccording to increasing intensity, just one panellist was removedfrom the panel following the criteria used for the selection (Spear-man’s ranked correlation coefficient was not significant).  3.1.2. Panel training  The objective colour of the training solutions, both commercialorange juices and their dilutions, were analysed by image analysis.The values of the coordinate  L ⁄ ranged from 64.70 to 72.03; hueranged from 78.5 (the most reddish OJ) to 92.8 (the most yellow)and the chroma ranged from 42.65 to 60.79. Fig. 5 shows the sam-ples in the CIELAB space and the coordinate  L ⁄ .Panellists were trained in different sessions until a consistentpanel of assessors was obtained. Fig. 6 shows the evolution inthe scores for the sample COJ I along the sessions. Standard devia-tions decreased fromthe first to the last session (from1.66 to 0.55and from 1.75 to 0.58 in hue and intensity, respectively). The errorbetween the last and the penultimate session was 6.6% in hue and4.5% in intensity (both cases less than 10%) in agreement with anincreasing uniformity of the panel. Finally, eight panellists wereselected to be part of the panel. Fig. 3.  Examples of the results for the Farnsworth-Munsell 100-Hue Test conducted in panellists with (a) average discrimination and (b) superior discrimination. R. Fernández-Vázquez et al./Food Quality and Preference 30 (2013) 320–327   323
Related Search
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!