Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Image processing analysis to track colour changes on apple and correlate to moisture content in drying stages

Image processing analysis to track colour changes on apple and correlate to moisture content in... Colour and moisture content are two most important attributes of the commercial food product. Estimation of moisture content is very important to know the storability of the food product. It also relates to the process of drying in a fruit or vegetable. Extra drying and shrinkage deteriorates the quality of food product. The goal of the experiment was to examine the changes in RGB values of an apple during drying at different temperatures. In this study, emphasis was given on how the colour changes when there is a significant change in the moisture content of the apple. Three randomly chosen varieties of apples were sliced to 8 mm thickness and dried in vacuum oven at 60°C, 70°C, and 80°C. The loss of moisture was recorded for every 30 min interval and corresponding digital images were taken to determine the change in RGB value. The images that were captured during the study was analysed in MATLAB image analysis computer software. The analysis of moisture content and average colour share with respect to time showed that average colour share value decreases with time at all three temperatures. More than 50 per cent of variation in moisture content was explained by average colour share. There is a significant linear relationship between moisture content and colour changes in RGB and can be used to predict the moisture content of apple during drying process. Key words: image processing; colour; moisture content; apple drying; temperature. influenced by the methods of drying, and the variables used ( Krokida Introduction and Maroulis, 1997). Seiiedlou et al., (2010) assessed the effects of dry- Image processing can be used to study the colour of fruits and veg- ing on quality characteristics like shrinkage and colour of dried apple etable and hence helps us to describe the quality of the material. It which were dried using a hot-air tray dryer. The changes in colour are is tough for naked human eyes to find out minute colour variations the result of drying along with certain enzymatic and non-enzymatic and identify the quality of an item. Computer imaging could be the reactions leading to browning (Vadivambal and Jayas, 2007). The best alternative for effective acquisition of information from colour changes in colour during drying are affected by different ongoing pro- images. Computer imaging involves digital colour camera, colour cedures: use of colour protection agent, temperature deviation, and monitor, and a simple software for analysis (Bora et al., 2015). intermittent drying (Krokida et. al., 2007). Lozano and Ibarz, (1997) Drying is a classical method of food preservation and an area studied the colour change in various fruits after heat treatments. where developments can be done via research (Velic et  al., 2004; Texture reveals the mechanical and microstructural properties of any Sacilik and Elicin, 2006). Fruits can be dried using different methods fruit, which plays an important role in defining the quality of any dried and instruments (oven and dryer). However, it is difficult to achieve the food (Martynenkoa and Janaszek, 2014). It would be of great impor- good quality of dried food products during the processing. The qual- tance to observe and control the texture development during drying. ity and physicochemical characteristics of the dried food products are © The Author(s) 2018. Published by Oxford University Press on behalf of Zhejiang University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/ licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 106 G. C. Bora et al., 2018, Vol. 2, No. 2 Apple is a low calorie fruit with high fibre and vitamin C. It might also ranges from 0 to 255, and the value close to 0 represents com- help us to reduce the effects of asthma, maintain the weight, and the pletely impure colour, whereas 255 represents completely pure col- nutrients present can reduce cholesterol levels. So, after processing our (Jensen, 2005). it is also necessary to retain the quality for nutritional values and Image processing has various applications in different fields. health benefits. There are several studies done previously on apples Recently, it is being used in different aspects of agriculture mainly using image processing to comment on the quality. For example, because it is a non-destructive method. Image processing involves Shahin et al. (2002) investigated two varieties of apples for separat- taking multiple photographs to measure different set of charac- ing damaged apples from the others. Vesali et  al. (2011) estimated teristics. It usually treats images as two-dimensional signals which the moisture content of apple with the help of image processing and generally involves three steps. First step is to take the digital image; simple weighing machine. Later, they also incorporated the neural then manipulating, processing, and analysing the image; and finally network for further analysis. Veraverbeke et al. (2003) evaluated the results based on the image analysis which may be an altered image. specific effect of the cuticle structure and cracks in the modelling of Image processing is an effective method used to identify the quality, moisture loss during long-term storage of apples. With long-term moisture content, volume, stages of ripening, any diseased condition, storage, there is reduction in moisture content of apples. The rate at yield, canopy, etc., of agricultural food products. Image processing which the moisture loss happens varies to different temperatures the technology has been found to be an effective technique and shown apples are subjected. improved accuracies in determining the vegetation indices, canopy Furthermore, recently researchers are finding the image process - measurement, and irrigated land mapping (Vibhute and Bodhe, ing method that is very efficient in determining other characteristics 2012). With the advancement of the technologies in the agricultural of various fruits and crops. As reduction in moisture content also area, image processing has become an easy, ecological friendly and causes reduction in weight, in turn reducing the economic value of cost effective method to study important parameters of agricultural apples. A study done by Mustafa et al. (2008) considered the quality products when compared with the conventional method. of the banana by determining the ripeness and size of the banana Changes in moisture content often induce changes in the colour using the image processing toolbox in MATLAB. McDonald and content of the food product. According to Jokic et al., rehydration Chen (1990) illustrated application of morphological image pro- rates and colour characteristic of apple samples are dependent on cessing with demonstration of three examples involving corn ker- differential drying conditions. The change in shape and volume and nel size discrimination, plant leaf identification, and texture analysis extra hardness in the produce cause bad impression on the cus- of marbling in beef longissimus dorsi muscle. Ribeiro et  al. (2005) tomer. Therefore, it causes changes in economic value of the food developed a computer-based image processing system to estimate the product and shifts the consumer demand of the product. Air dry- weed pressure. Mateos et al. (2014) used image processing in irriga- ing tends to increase the degree shrinkage and destroy the cellu- tion management applications. lar structure (Sturm et al., 2012). Change in moisture content also Usually colour images are displayed in three primary colour com- causes alteration in the weight of the food product. Dehydration binations Red–Green–Blue, which is based on additive colour theory. and evaporation of surface water bring changes in the colour of the Additive theory explains after effects of light mixing rather than material. Colour change such as browning or yellowing in fruits when pigments are mixed as in subtractive theory (Jensen, 2005). refers to represent the deteriorating quality of that fruit. While Information can be depicted in terms of chromaticity coordinates, measuring the quality of any commercial fruit, colour and moisture which is used to specify colours. The coordinates here represent the content are two significant parameters. Apple, scientifically known relative amount of each primary colour as given in Equations (1) as Malus domestica, is one of the delicious and popular fruit con- through (3). The sum of primary colour is always one as shown in taining some essential nutrients good for health. There are lots of the following equation: health benefits of apple fruit. Apple contains different phytochemi - cals like quercetin, epicatechin, and procyanidin that are beneficial x = , (1) for human health. Apple also contains soluble and insoluble fibre RG ++ B that helps in digestion process. Food drying is one of the traditional G and oldest methods of food preservation. For commercial purpose, y = , (2) RG ++ B it is necessary that it can be stored for few months or so. Preserving apple is essential because apples are not only eaten raw but used to make desserts, jams, candies, cakes, etc. Dried fruit products z = , (3) RG ++ B like apple is good for widening product availability and to diver- sify markets (Contreras et al., 2008). Drying removes the moisture x ++ y z = 1, (4) content which in turn reduces the bacterial and yeast activity and preserves the food for a long time. Also reduction in weight ease where R is red, B is blue, and G is green and they represent the handling and storage. Fruits can be dried in sun and oven both with amount of red, green, and blue needed to form any definite colour, necessary combination of time and temperature. Moisture is evapo- and x, y, and z are normalized colour components known as trichro- rated during drying in warm temperatures. Low humidity enhances matic coefficients ( Jensen, 2005). the movement of moisture to external ambient, whereas air cur- Another colour coordinate is Hue-Saturation-Intensity (H-I-S) rent speeds up the process by replacing the surrounding moist air. which is based on hypothetical colour sphere. Hue is the attribute Dried fruits are supposed to be appealing, long lasting, tasty, and of colour perception through which one can identify any specific nutritious. Therefore, it is important to study the changes of apple colour. The value of hue begins with 0 and increases counter-clock- textures and colour at different temperatures. wise before finishing to 255. Intensity does not associate with any The main objective of the study is to investigate the drying of colour, it is just the relative of darkness and varies from dark (0) apple at different stages and temperatures to evaluate the changes in to white (255). Saturation is simply the purity of colour, the value colour and correlate it to the moisture content. Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 Image Processing Analysis, 2018, Vol. 2, No. 2 107 drying condition for all samples. Stopwatch available in IPhone was Material and Methods used for recording the drying time. D3100 SLR 14.2 MP Nikon digi- Three varieties of apples, Red delicious, Granny smith, and Gold tal camera was used for taking digital images of the sample. Digital delicious cultivar used in the study, were obtained from local store weighing machine was used for the measurement of sample weights. at Fargo. The three varieties were named as A, B, and C, respectively, A pink board with two colour circles, green and yellow as shown in for simplicity of analysis. Samples were bought and stored in a lab Figure 2, was used for taking images so that it can be calibrated for refrigerator. Three uniform size apples for each variety were selected the colour values with circles. and washed with tap water and wiped with tissue paper. Each apple was cut vertical to the axis into 8 slices of 8  mm thickness with the help of a locally available slicer shown in Figure  1. Only three wholesome slices from the middle of almost equal sizes were selected for drying. First three and last two slices were discarded for further processing. Then the seeds were manually removed with a knife. The experimental apparatus, vacuum oven (Model LBB1-69A-1, Despatch, Minneapolis, MN, USA), capable of maximum tempera- ture of 240°C (400 F) with 2.4 KW heater, was used for drying the apple slices. Vacuum oven is generally used to measure the amount of water present in a material. Vacuum drying reduces the moisture content in an object with air drier. Vacuum drying involves reduced pressure environment which decreases the heat needed for speedy drying. This process requires less energy and is less damaging. Vacuum oven consists of two chambers with meshed tray–like permanent structures, allowing unrestricted air circulation. The air flow and temperature can be adjusted. Apple slices were uniformly distributed in thin layer in both the trays. The Figure 3. Relationship between dry basis moisture content and drying time at 60°C. samples were shuffled at each repetition to maintain the uniform Figure 4. Relationship between dry basis moisture content and drying time at 70°C. Figure 1. Eight millimetre thick apple slicer. Figure 5. Relationship between dry basis moisture content and drying time Figure 2. Pink board with two colour circles: green and yellow. at 70°C. Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 108 G. C. Bora et al., 2018, Vol. 2, No. 2 Table 1. Time, temperature, percent moisture content, and colour changes in RGB for three varieties of apple. Time Temperature (min) 60°C 70°C 80°C A B C A B C A B C %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E RGB RGB RGB RGB RGB RGB RGB RGB RGB 0 416.42 80.75 396.52 65.25 467.36 71.28 480.65 35.58 399.31 38.25 479.35 32.32 453.51 59.89 487.68 27.16 471.01 30.01 30 362.71 25.92 350.30 22.86 407.54 35.14 410.37 12.05 344.13 14.99 418.26 11.55 367.82 40.73 397.21 16.41 384.02 12.49 60 308.70 27.47 291.94 20.34 338.89 33.30 343.67 18.45 284.02 12.69 359.78 16.12 295.08 46.63 314.44 26.13 311.54 24.94 90 267.46 26.87 246.18 19.71 285.80 34.07 285.68 17.52 230.54 12.98 288.29 15.49 228.69 49.65 239.25 28.88 242.40 22.04 120 225.34 26.85 211.43 24.28 236.03 31.77 227.35 21.69 185.59 15.87 222.92 19.35 170.79 31.09 170.47 33.39 183.55 17.37 150 187.53 24.39 176.92 16.79 198.91 19.10 176.07 25.45 142.65 23.57 164.22 22.45 118.46 17.25 112.53 14.66 128.10 4.15 180 153.97 20.85 143.60 24.02 157.92 23.75 123.93 21.32 99.56 15.09 112.90 11.76 71.23 17.00 60.07 25.99 74.81 14.22 210 122.45 17.89 110.76 15.98 120.75 19.49 83.57 14.18 64.94 14.04 69.89 9.12 32.88 13.35 38.45 19.10 19.81 8.20 240 95.06 15.75 84.14 13.85 88.03 8.75 51.41 11.48 39.55 13.17 35.22 4.15 10.96 8.15 9.28 20.95 6.63 4.48 270 66.56 12.80 59.03 9.21 55.92 7.90 26.61 10.33 20.96 7.09 13.56 9.30 2.39 10.19 1.94 15.89 1.97 2.74 300 43.56 12.72 40.52 5.83 32.50 7.07 12.67 7.43 9.54 9.46 4.75 3.24 0.34 10.46 0.18 9.75 0.37 4.41 330 26.97 13.42 25.30 14.46 15.31 9.09 4.01 7.27 3.50 4.34 1.82 4.05 0.00 0.00 0.00 0.00 0.00 0.00 360 14.07 24.92 13.92 17.46 6.16 15.16 0.65 4.31 0.95 5.32 0.57 4.88 390 7.20 7.23 7.44 6.24 3.43 12.29 0.00 0.00 0.00 0.00 0.00 0.00 420 2.96 4.78 3.07 10.43 1.71 9.41 450 1.50 7.18 1.49 7.00 0.89 13.14 480 0.05 6.17 0.29 9.79 0.36 13.59 510 0.00 0.00 0.00 0.00 0.00 0.00 Figure 6. Relationship between average colour share and drying time at 60°C. Figure 8. Relationship between average colour share and drying time at 80°C. so as to get the drying curves. Also digital images were taken for each corresponding record of moisture loss to determine the change in RGB values. Drying was continued until there was no change observed in the sample weight. The experiment was replicated for 70°C and 80°C. Dry basis moisture content was determined using the following equation (Wilhelm et al., 2004): M = , (5) where M = decimal moisture content dry basis (db), md = mass of dry matter in the product, and mw = mass of water in the product. Percent moisture content is found by multiplying the decimal mois- ture content by 100. Dry basis is generally used to measure the moisture Figure 7. Relationship between average colour share and drying time at 70°C. content of any material during drying process. Dry basis moisture con- tent can be defined as the amount of water per unit mass of dry solids in Drying of apple slices was conducted in three different tempera- the sample. The moisture content for high moisture materials like fruits tures of 60°C, 70°C, and 80°C. Apple slices were put on the mesh and vegetables can go up to 900 per cent on a dry basis. tray sparsely ensuring adequate air circulation. The oven tempera- Digital images were processed by MATLAB software in RGB col- ture was set to 60°C. Samples were taken out of the oven and mois- our model to evaluate colour changes in each different stage of drying. ture loss was recorded for every 30  min interval for each sample, Program calculated the average percentage of red (R), green (G), and Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 Image Processing Analysis, 2018, Vol. 2, No. 2 109 blue (B) colours on sample area. Average RGB values obtained from in moisture content as a function of drying time for the three dif- MATLAB were exported directly to excel sheet for ease of data analysis. ferent varieties (A, B, and C). Table 1 shows the percent moisture For decisive analysis of colour changes in RGB model, the fol- content, colour changes in RGB model with respect to time and lowing formula was used: Average share of each colour, temperature. From the graphs in Figures 3–5, for all three varieties, the dry 22 2 basis moisture content has negative linear relationship with time   ΔΔ ER = + ΔΔ GB + (6) () () () RGB   of drying, that is, the moisture content decreases continuously by the time, which is a normal phenomenon. There is also direct rela- where ∆R, ∆G, and ∆B are differences between colour values of fresh tionship between complete drying time and the temperature; with samples and dried samples, and ΔE are the colour changes in RGB increase in temperature, there is reduction in drying time. It showed RGB model. that the time taken to completely dry the apple was the same in SAS software was used for statistical analysis. PROC REG pro- all the three varieties: 330  min for 80°C, 390  min for 70°C, and cedure was employed for model statements. Analysis of variance 510 min for 60°C. (ANOVA) table, Root MSE, and R-square values were obtained. SAS Proc Reg model was used for the linear regression analysis of Data were analysed at 95% confidence level. the moisture content response to colour changes. From the graphs in Figures 6–8, we can see the change in colour (Average colour share, ∆E ) for all three varieties with respect to time. We can see that RGB Results and Analysis average colour share value decreases with increase in time in all the The drying of three different varieties of apple slices of thick- three temperatures, though the changes are not smooth and there are ness 8  mm at three temperatures of 60°C, 70°C, and 80°C and some fluctuations. It is interesting to note that at all three tempera - their colour changes were studied. For the simplicity, interpret- tures, there is a rapid drop in the RGB colour share value at the first ability, and common acceptance of linear model, fit for linear point of data collection, which is at 30 min. There is a sharp increase model was explored. Figures 3, 4, and 5 represent the variation in the value at around 350 min for 60°C in all three varieties. The Figure 9. The relationship between average colour share and moisture content for variety A at different temperatures (a through b). Figure 10. The relationship between average colour share and moisture content for variety B at different temperatures (a through b). Figure 11. The relationship between average colour share and moisture content for variety C at different temperatures (a through b). Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 110 G. C. Bora et al., 2018, Vol. 2, No. 2 Garcia-Mateos, G., Hernandez-Hernandez, J. L., Escarabajal-Henarejos, D., sharp rise is 150 min for 70°C and 100 min for 80°C, respectively. Jaen-Terrones, S., Molina-Martinez, J. M. (2014). Study and comparison Figures 9–11 illustrate the correlation between the moisture content of color models for automatic image analysis in irrigation management and average colour share for all treatment combinations of tempera- applications. Agricultural Water Management, 151: 158–166 tures and varieties. Jensen, R. J., 2005. Introductory Digital Image Processing. Prentice Hall Series The linear relationship between moisture content and average in Geographic Information Sciences. Third Edition. colour share is statistically significant at 5 per cent with adjusted Jokic, S., Lukinac, J., Velic, D., Bilic, M., Magdic, D., Planinic, M. Optimization coefficient of determination: R-square value of 69 per cent for 60°C, of Drying Parameters and Color Changes of Pretreated Organic Apple 54 per cent for 70°C, and 45 per cent for 80°C temperatures, respec- Slices. Department of Process Engineering, Faculty of Food Technology, tively. For natural products like fruits and vegetables, it is likely to University J.J. Strossmayer of Osijek, Croatia. have a lower coefficient of determination. The linear models, derived Krokida, M. K., Maroulis, Z. B. (1997). Effect of drying method on shrinkage and porosity. Drying Technology-An International Journal, 15: 1145–1155 from statistical analysis using SAS Proc Reg, are given from the fol- Krokida, K. M., Kiranoudis, T. C., Maroulis, B. Z., Marinos-kouris, D., (2007). lowing equations: Drying related properties of Apple. Drying Technology-An International Journal, 18-6: 1570–1582. For600 °= CColor Share .. 0953 Moisture Content + 6 812, (7) Lozano, J. E., Ibarz, A. (1997). Colour changes in concentrated fruit pulp during heating at high temperatures. Journal of Food Engineering, 31: For700 °= CColor Share .. 0432 Moisture Content + 6 881, (8) 365–373. Martynenko, A., Janaszek, A. M. (2014). Texture changes during drying of apple slices. Drying Technology-An International Journal, 32–5: 567–577. For800 °= CColor Share .. 0611 Moisture Content + 9 924. (9) McDonald, T., Chen, Y. (1990). Application of morphological image process- ing in agriculture. Transactions of the ASAE, 33: 1345–1352 Mustafa, N. B. A., et al. (2008). Image Processing of an Agriculture Produce: Conclusion Determination of Size and Ripeness of a Banana. International Symposium on Information Technology, Kuala Lumpur, 1–7. The project was undertaken to study the changes in colour and Ribeiro, A., Fernandez-Quintanilla, C., Barroso, J., Garcia-Alegre, M. C., moisture content of three different varieties of apple at multiple tem- Stafford, J. V. (2005). Development of an image analysis system for peratures with time. The moisture content and average colour share estimation of weed pressure. In: Precision Agriculture’05. 5th European value decreased with increase in time. For higher temperature, there Conference on Precision Agriculture, Uppsala, Swede, Wageningen is a lower drying time, since the rate of evaporation of the product Academic Publishers, 169–174. increases with the temperature. It was also found that in the drying Sacilik, K., Elicin, A. K. (2006). The thin layer drying characteristics of organic process of apple slices that the colour share value dropped signifi - apple slices. Journal of Food Engineering, 73: 281–289. cantly within 30 min, but the moisture content decreased gradually Seiiedlou, S. A.  D. E.  G. H., Ghasemzadeh, H. R., Hamdami, N., Talati, F. for all the varieties in all temperatures. Good coefficient of determi - A. R. A. M. A. R. Z., Moghaddam, M. (2010). Convective drying of apple: nation was established with significant linear relationship at lower mathematical modeling and determination of some quality parameters. International Journal of Agriculture and Biology, 12: 171–178. temperature values of 60°C in colour share values. With increase in Shahin M. A., Tollner E. W., McClendon R. W., Arabnia H. R. (2002). Apple temperature, decrease in the coefficient of determination was noted. classification based on surface bruises using image processing and neural The adjusted coefficient of determination was found to be around networks. Transactions of the ASAE, 45: 1619–1627. 69 per cent for 60°C, 54 per cent for 70°C, and 45 per cent for 80°C Sturm, B., Hofacker C. W., Hensel O. (2012). Optimizing the drying parame- temperatures, respectively, for the linear model of moisture content ters for hot-air–dried apples. Drying Technology-An International Journal, and the colour share value, as more than 50 per cent of change in 30-14: 1570–1582. moisture content was validated by average colour share, which is Vadivambal, R., Jayas, D. S. (2007). Changes in quality of microwave-treated reasonable for natural products. Linear relationship is established as agricultural products-a review. Biosystems Engineering, 98: 1–16. the best fit to identify the change in moisture content through aver - Velic, D., Planinic, M., Tomas, S., Bilic, M. (2004). Influence of airflow veloc - age colour share values. For further research and better results, the ity on kinetics of convection apple drying. Journal of Food Engineering, 64: 97–102. number of samples can be increased and digital information can be Veraverbeke E. A., Verboven P., Van Oostveldt P., Nicolaï B. M. (2003) captured at controlled ambient conditions. Prediction of moisture loss across the cuticle of apple [Malus sylves- trissubsp. mitis (Wallr.)] during storage: part 2: model simulations References and practical applications. Postharvest Biology and Technology, 30: 75–88. Bora, C. G., Lin, D., Bhattacharya, P., Bali, K. S., Pathak, R. (2015). Application Vesali, F., Gharibkhani, M., Komarizadeh, M. H. (2011). An approach to esti- of bio-image analysis for classification of different ripening stages of mate moisture content of apple with image processing method. Australian banana. Journal of Agricultural Science, 7-2: 1916–9752. Journal of Crop Science, 5: 111–115 Contreras, C., Martin-Esparza, M. E., Chiralt, A., Martinez-Navarrete, N. Vibhute, A., Bodhe, S. K. (2012). Application of image processing in agri- (2008). Influence of microwave application on convective drying. Effects culture: a survey. International Journal of Computer Application, 52: on drying kinetics, and optical and mechanical properties of apple and 34–40 strawberry. Journal of Food Engineering, 88: 55–64. Wilhelm, L. R., Suter, D. A., Brusewitz, G. H. (2004). Chapter 10: Drying and Du, C. J., Sun, D. W. (2004). Recent developments in the applications of image dehydration. In: Food and Process Engineering Technology. ASAE, ST. processing techniques for food quality evaluation. Trends in Food Science Joseph, Michigan, pp. 259–284. & Technology, 15: 230–249. Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Food Quality and Safety Oxford University Press

Image processing analysis to track colour changes on apple and correlate to moisture content in drying stages

Food Quality and Safety , Volume Advance Article (2) – May 12, 2018

Loading next page...
 
/lp/oxford-university-press/image-processing-analysis-to-track-colour-changes-on-apple-and-eHo7G8qq0O

References (23)

Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of Zhejiang University Press.
ISSN
2399-1399
eISSN
2399-1402
DOI
10.1093/fqsafe/fyy003
Publisher site
See Article on Publisher Site

Abstract

Colour and moisture content are two most important attributes of the commercial food product. Estimation of moisture content is very important to know the storability of the food product. It also relates to the process of drying in a fruit or vegetable. Extra drying and shrinkage deteriorates the quality of food product. The goal of the experiment was to examine the changes in RGB values of an apple during drying at different temperatures. In this study, emphasis was given on how the colour changes when there is a significant change in the moisture content of the apple. Three randomly chosen varieties of apples were sliced to 8 mm thickness and dried in vacuum oven at 60°C, 70°C, and 80°C. The loss of moisture was recorded for every 30 min interval and corresponding digital images were taken to determine the change in RGB value. The images that were captured during the study was analysed in MATLAB image analysis computer software. The analysis of moisture content and average colour share with respect to time showed that average colour share value decreases with time at all three temperatures. More than 50 per cent of variation in moisture content was explained by average colour share. There is a significant linear relationship between moisture content and colour changes in RGB and can be used to predict the moisture content of apple during drying process. Key words: image processing; colour; moisture content; apple drying; temperature. influenced by the methods of drying, and the variables used ( Krokida Introduction and Maroulis, 1997). Seiiedlou et al., (2010) assessed the effects of dry- Image processing can be used to study the colour of fruits and veg- ing on quality characteristics like shrinkage and colour of dried apple etable and hence helps us to describe the quality of the material. It which were dried using a hot-air tray dryer. The changes in colour are is tough for naked human eyes to find out minute colour variations the result of drying along with certain enzymatic and non-enzymatic and identify the quality of an item. Computer imaging could be the reactions leading to browning (Vadivambal and Jayas, 2007). The best alternative for effective acquisition of information from colour changes in colour during drying are affected by different ongoing pro- images. Computer imaging involves digital colour camera, colour cedures: use of colour protection agent, temperature deviation, and monitor, and a simple software for analysis (Bora et al., 2015). intermittent drying (Krokida et. al., 2007). Lozano and Ibarz, (1997) Drying is a classical method of food preservation and an area studied the colour change in various fruits after heat treatments. where developments can be done via research (Velic et  al., 2004; Texture reveals the mechanical and microstructural properties of any Sacilik and Elicin, 2006). Fruits can be dried using different methods fruit, which plays an important role in defining the quality of any dried and instruments (oven and dryer). However, it is difficult to achieve the food (Martynenkoa and Janaszek, 2014). It would be of great impor- good quality of dried food products during the processing. The qual- tance to observe and control the texture development during drying. ity and physicochemical characteristics of the dried food products are © The Author(s) 2018. Published by Oxford University Press on behalf of Zhejiang University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/ licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 106 G. C. Bora et al., 2018, Vol. 2, No. 2 Apple is a low calorie fruit with high fibre and vitamin C. It might also ranges from 0 to 255, and the value close to 0 represents com- help us to reduce the effects of asthma, maintain the weight, and the pletely impure colour, whereas 255 represents completely pure col- nutrients present can reduce cholesterol levels. So, after processing our (Jensen, 2005). it is also necessary to retain the quality for nutritional values and Image processing has various applications in different fields. health benefits. There are several studies done previously on apples Recently, it is being used in different aspects of agriculture mainly using image processing to comment on the quality. For example, because it is a non-destructive method. Image processing involves Shahin et al. (2002) investigated two varieties of apples for separat- taking multiple photographs to measure different set of charac- ing damaged apples from the others. Vesali et  al. (2011) estimated teristics. It usually treats images as two-dimensional signals which the moisture content of apple with the help of image processing and generally involves three steps. First step is to take the digital image; simple weighing machine. Later, they also incorporated the neural then manipulating, processing, and analysing the image; and finally network for further analysis. Veraverbeke et al. (2003) evaluated the results based on the image analysis which may be an altered image. specific effect of the cuticle structure and cracks in the modelling of Image processing is an effective method used to identify the quality, moisture loss during long-term storage of apples. With long-term moisture content, volume, stages of ripening, any diseased condition, storage, there is reduction in moisture content of apples. The rate at yield, canopy, etc., of agricultural food products. Image processing which the moisture loss happens varies to different temperatures the technology has been found to be an effective technique and shown apples are subjected. improved accuracies in determining the vegetation indices, canopy Furthermore, recently researchers are finding the image process - measurement, and irrigated land mapping (Vibhute and Bodhe, ing method that is very efficient in determining other characteristics 2012). With the advancement of the technologies in the agricultural of various fruits and crops. As reduction in moisture content also area, image processing has become an easy, ecological friendly and causes reduction in weight, in turn reducing the economic value of cost effective method to study important parameters of agricultural apples. A study done by Mustafa et al. (2008) considered the quality products when compared with the conventional method. of the banana by determining the ripeness and size of the banana Changes in moisture content often induce changes in the colour using the image processing toolbox in MATLAB. McDonald and content of the food product. According to Jokic et al., rehydration Chen (1990) illustrated application of morphological image pro- rates and colour characteristic of apple samples are dependent on cessing with demonstration of three examples involving corn ker- differential drying conditions. The change in shape and volume and nel size discrimination, plant leaf identification, and texture analysis extra hardness in the produce cause bad impression on the cus- of marbling in beef longissimus dorsi muscle. Ribeiro et  al. (2005) tomer. Therefore, it causes changes in economic value of the food developed a computer-based image processing system to estimate the product and shifts the consumer demand of the product. Air dry- weed pressure. Mateos et al. (2014) used image processing in irriga- ing tends to increase the degree shrinkage and destroy the cellu- tion management applications. lar structure (Sturm et al., 2012). Change in moisture content also Usually colour images are displayed in three primary colour com- causes alteration in the weight of the food product. Dehydration binations Red–Green–Blue, which is based on additive colour theory. and evaporation of surface water bring changes in the colour of the Additive theory explains after effects of light mixing rather than material. Colour change such as browning or yellowing in fruits when pigments are mixed as in subtractive theory (Jensen, 2005). refers to represent the deteriorating quality of that fruit. While Information can be depicted in terms of chromaticity coordinates, measuring the quality of any commercial fruit, colour and moisture which is used to specify colours. The coordinates here represent the content are two significant parameters. Apple, scientifically known relative amount of each primary colour as given in Equations (1) as Malus domestica, is one of the delicious and popular fruit con- through (3). The sum of primary colour is always one as shown in taining some essential nutrients good for health. There are lots of the following equation: health benefits of apple fruit. Apple contains different phytochemi - cals like quercetin, epicatechin, and procyanidin that are beneficial x = , (1) for human health. Apple also contains soluble and insoluble fibre RG ++ B that helps in digestion process. Food drying is one of the traditional G and oldest methods of food preservation. For commercial purpose, y = , (2) RG ++ B it is necessary that it can be stored for few months or so. Preserving apple is essential because apples are not only eaten raw but used to make desserts, jams, candies, cakes, etc. Dried fruit products z = , (3) RG ++ B like apple is good for widening product availability and to diver- sify markets (Contreras et al., 2008). Drying removes the moisture x ++ y z = 1, (4) content which in turn reduces the bacterial and yeast activity and preserves the food for a long time. Also reduction in weight ease where R is red, B is blue, and G is green and they represent the handling and storage. Fruits can be dried in sun and oven both with amount of red, green, and blue needed to form any definite colour, necessary combination of time and temperature. Moisture is evapo- and x, y, and z are normalized colour components known as trichro- rated during drying in warm temperatures. Low humidity enhances matic coefficients ( Jensen, 2005). the movement of moisture to external ambient, whereas air cur- Another colour coordinate is Hue-Saturation-Intensity (H-I-S) rent speeds up the process by replacing the surrounding moist air. which is based on hypothetical colour sphere. Hue is the attribute Dried fruits are supposed to be appealing, long lasting, tasty, and of colour perception through which one can identify any specific nutritious. Therefore, it is important to study the changes of apple colour. The value of hue begins with 0 and increases counter-clock- textures and colour at different temperatures. wise before finishing to 255. Intensity does not associate with any The main objective of the study is to investigate the drying of colour, it is just the relative of darkness and varies from dark (0) apple at different stages and temperatures to evaluate the changes in to white (255). Saturation is simply the purity of colour, the value colour and correlate it to the moisture content. Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 Image Processing Analysis, 2018, Vol. 2, No. 2 107 drying condition for all samples. Stopwatch available in IPhone was Material and Methods used for recording the drying time. D3100 SLR 14.2 MP Nikon digi- Three varieties of apples, Red delicious, Granny smith, and Gold tal camera was used for taking digital images of the sample. Digital delicious cultivar used in the study, were obtained from local store weighing machine was used for the measurement of sample weights. at Fargo. The three varieties were named as A, B, and C, respectively, A pink board with two colour circles, green and yellow as shown in for simplicity of analysis. Samples were bought and stored in a lab Figure 2, was used for taking images so that it can be calibrated for refrigerator. Three uniform size apples for each variety were selected the colour values with circles. and washed with tap water and wiped with tissue paper. Each apple was cut vertical to the axis into 8 slices of 8  mm thickness with the help of a locally available slicer shown in Figure  1. Only three wholesome slices from the middle of almost equal sizes were selected for drying. First three and last two slices were discarded for further processing. Then the seeds were manually removed with a knife. The experimental apparatus, vacuum oven (Model LBB1-69A-1, Despatch, Minneapolis, MN, USA), capable of maximum tempera- ture of 240°C (400 F) with 2.4 KW heater, was used for drying the apple slices. Vacuum oven is generally used to measure the amount of water present in a material. Vacuum drying reduces the moisture content in an object with air drier. Vacuum drying involves reduced pressure environment which decreases the heat needed for speedy drying. This process requires less energy and is less damaging. Vacuum oven consists of two chambers with meshed tray–like permanent structures, allowing unrestricted air circulation. The air flow and temperature can be adjusted. Apple slices were uniformly distributed in thin layer in both the trays. The Figure 3. Relationship between dry basis moisture content and drying time at 60°C. samples were shuffled at each repetition to maintain the uniform Figure 4. Relationship between dry basis moisture content and drying time at 70°C. Figure 1. Eight millimetre thick apple slicer. Figure 5. Relationship between dry basis moisture content and drying time Figure 2. Pink board with two colour circles: green and yellow. at 70°C. Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 108 G. C. Bora et al., 2018, Vol. 2, No. 2 Table 1. Time, temperature, percent moisture content, and colour changes in RGB for three varieties of apple. Time Temperature (min) 60°C 70°C 80°C A B C A B C A B C %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E %MC ∆E RGB RGB RGB RGB RGB RGB RGB RGB RGB 0 416.42 80.75 396.52 65.25 467.36 71.28 480.65 35.58 399.31 38.25 479.35 32.32 453.51 59.89 487.68 27.16 471.01 30.01 30 362.71 25.92 350.30 22.86 407.54 35.14 410.37 12.05 344.13 14.99 418.26 11.55 367.82 40.73 397.21 16.41 384.02 12.49 60 308.70 27.47 291.94 20.34 338.89 33.30 343.67 18.45 284.02 12.69 359.78 16.12 295.08 46.63 314.44 26.13 311.54 24.94 90 267.46 26.87 246.18 19.71 285.80 34.07 285.68 17.52 230.54 12.98 288.29 15.49 228.69 49.65 239.25 28.88 242.40 22.04 120 225.34 26.85 211.43 24.28 236.03 31.77 227.35 21.69 185.59 15.87 222.92 19.35 170.79 31.09 170.47 33.39 183.55 17.37 150 187.53 24.39 176.92 16.79 198.91 19.10 176.07 25.45 142.65 23.57 164.22 22.45 118.46 17.25 112.53 14.66 128.10 4.15 180 153.97 20.85 143.60 24.02 157.92 23.75 123.93 21.32 99.56 15.09 112.90 11.76 71.23 17.00 60.07 25.99 74.81 14.22 210 122.45 17.89 110.76 15.98 120.75 19.49 83.57 14.18 64.94 14.04 69.89 9.12 32.88 13.35 38.45 19.10 19.81 8.20 240 95.06 15.75 84.14 13.85 88.03 8.75 51.41 11.48 39.55 13.17 35.22 4.15 10.96 8.15 9.28 20.95 6.63 4.48 270 66.56 12.80 59.03 9.21 55.92 7.90 26.61 10.33 20.96 7.09 13.56 9.30 2.39 10.19 1.94 15.89 1.97 2.74 300 43.56 12.72 40.52 5.83 32.50 7.07 12.67 7.43 9.54 9.46 4.75 3.24 0.34 10.46 0.18 9.75 0.37 4.41 330 26.97 13.42 25.30 14.46 15.31 9.09 4.01 7.27 3.50 4.34 1.82 4.05 0.00 0.00 0.00 0.00 0.00 0.00 360 14.07 24.92 13.92 17.46 6.16 15.16 0.65 4.31 0.95 5.32 0.57 4.88 390 7.20 7.23 7.44 6.24 3.43 12.29 0.00 0.00 0.00 0.00 0.00 0.00 420 2.96 4.78 3.07 10.43 1.71 9.41 450 1.50 7.18 1.49 7.00 0.89 13.14 480 0.05 6.17 0.29 9.79 0.36 13.59 510 0.00 0.00 0.00 0.00 0.00 0.00 Figure 6. Relationship between average colour share and drying time at 60°C. Figure 8. Relationship between average colour share and drying time at 80°C. so as to get the drying curves. Also digital images were taken for each corresponding record of moisture loss to determine the change in RGB values. Drying was continued until there was no change observed in the sample weight. The experiment was replicated for 70°C and 80°C. Dry basis moisture content was determined using the following equation (Wilhelm et al., 2004): M = , (5) where M = decimal moisture content dry basis (db), md = mass of dry matter in the product, and mw = mass of water in the product. Percent moisture content is found by multiplying the decimal mois- ture content by 100. Dry basis is generally used to measure the moisture Figure 7. Relationship between average colour share and drying time at 70°C. content of any material during drying process. Dry basis moisture con- tent can be defined as the amount of water per unit mass of dry solids in Drying of apple slices was conducted in three different tempera- the sample. The moisture content for high moisture materials like fruits tures of 60°C, 70°C, and 80°C. Apple slices were put on the mesh and vegetables can go up to 900 per cent on a dry basis. tray sparsely ensuring adequate air circulation. The oven tempera- Digital images were processed by MATLAB software in RGB col- ture was set to 60°C. Samples were taken out of the oven and mois- our model to evaluate colour changes in each different stage of drying. ture loss was recorded for every 30  min interval for each sample, Program calculated the average percentage of red (R), green (G), and Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 Image Processing Analysis, 2018, Vol. 2, No. 2 109 blue (B) colours on sample area. Average RGB values obtained from in moisture content as a function of drying time for the three dif- MATLAB were exported directly to excel sheet for ease of data analysis. ferent varieties (A, B, and C). Table 1 shows the percent moisture For decisive analysis of colour changes in RGB model, the fol- content, colour changes in RGB model with respect to time and lowing formula was used: Average share of each colour, temperature. From the graphs in Figures 3–5, for all three varieties, the dry 22 2 basis moisture content has negative linear relationship with time   ΔΔ ER = + ΔΔ GB + (6) () () () RGB   of drying, that is, the moisture content decreases continuously by the time, which is a normal phenomenon. There is also direct rela- where ∆R, ∆G, and ∆B are differences between colour values of fresh tionship between complete drying time and the temperature; with samples and dried samples, and ΔE are the colour changes in RGB increase in temperature, there is reduction in drying time. It showed RGB model. that the time taken to completely dry the apple was the same in SAS software was used for statistical analysis. PROC REG pro- all the three varieties: 330  min for 80°C, 390  min for 70°C, and cedure was employed for model statements. Analysis of variance 510 min for 60°C. (ANOVA) table, Root MSE, and R-square values were obtained. SAS Proc Reg model was used for the linear regression analysis of Data were analysed at 95% confidence level. the moisture content response to colour changes. From the graphs in Figures 6–8, we can see the change in colour (Average colour share, ∆E ) for all three varieties with respect to time. We can see that RGB Results and Analysis average colour share value decreases with increase in time in all the The drying of three different varieties of apple slices of thick- three temperatures, though the changes are not smooth and there are ness 8  mm at three temperatures of 60°C, 70°C, and 80°C and some fluctuations. It is interesting to note that at all three tempera - their colour changes were studied. For the simplicity, interpret- tures, there is a rapid drop in the RGB colour share value at the first ability, and common acceptance of linear model, fit for linear point of data collection, which is at 30 min. There is a sharp increase model was explored. Figures 3, 4, and 5 represent the variation in the value at around 350 min for 60°C in all three varieties. The Figure 9. The relationship between average colour share and moisture content for variety A at different temperatures (a through b). Figure 10. The relationship between average colour share and moisture content for variety B at different temperatures (a through b). Figure 11. The relationship between average colour share and moisture content for variety C at different temperatures (a through b). Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018 110 G. C. Bora et al., 2018, Vol. 2, No. 2 Garcia-Mateos, G., Hernandez-Hernandez, J. L., Escarabajal-Henarejos, D., sharp rise is 150 min for 70°C and 100 min for 80°C, respectively. Jaen-Terrones, S., Molina-Martinez, J. M. (2014). Study and comparison Figures 9–11 illustrate the correlation between the moisture content of color models for automatic image analysis in irrigation management and average colour share for all treatment combinations of tempera- applications. Agricultural Water Management, 151: 158–166 tures and varieties. Jensen, R. J., 2005. Introductory Digital Image Processing. Prentice Hall Series The linear relationship between moisture content and average in Geographic Information Sciences. Third Edition. colour share is statistically significant at 5 per cent with adjusted Jokic, S., Lukinac, J., Velic, D., Bilic, M., Magdic, D., Planinic, M. Optimization coefficient of determination: R-square value of 69 per cent for 60°C, of Drying Parameters and Color Changes of Pretreated Organic Apple 54 per cent for 70°C, and 45 per cent for 80°C temperatures, respec- Slices. Department of Process Engineering, Faculty of Food Technology, tively. For natural products like fruits and vegetables, it is likely to University J.J. Strossmayer of Osijek, Croatia. have a lower coefficient of determination. The linear models, derived Krokida, M. K., Maroulis, Z. B. (1997). Effect of drying method on shrinkage and porosity. Drying Technology-An International Journal, 15: 1145–1155 from statistical analysis using SAS Proc Reg, are given from the fol- Krokida, K. M., Kiranoudis, T. C., Maroulis, B. Z., Marinos-kouris, D., (2007). lowing equations: Drying related properties of Apple. Drying Technology-An International Journal, 18-6: 1570–1582. For600 °= CColor Share .. 0953 Moisture Content + 6 812, (7) Lozano, J. E., Ibarz, A. (1997). Colour changes in concentrated fruit pulp during heating at high temperatures. Journal of Food Engineering, 31: For700 °= CColor Share .. 0432 Moisture Content + 6 881, (8) 365–373. Martynenko, A., Janaszek, A. M. (2014). Texture changes during drying of apple slices. Drying Technology-An International Journal, 32–5: 567–577. For800 °= CColor Share .. 0611 Moisture Content + 9 924. (9) McDonald, T., Chen, Y. (1990). Application of morphological image process- ing in agriculture. Transactions of the ASAE, 33: 1345–1352 Mustafa, N. B. A., et al. (2008). Image Processing of an Agriculture Produce: Conclusion Determination of Size and Ripeness of a Banana. International Symposium on Information Technology, Kuala Lumpur, 1–7. The project was undertaken to study the changes in colour and Ribeiro, A., Fernandez-Quintanilla, C., Barroso, J., Garcia-Alegre, M. C., moisture content of three different varieties of apple at multiple tem- Stafford, J. V. (2005). Development of an image analysis system for peratures with time. The moisture content and average colour share estimation of weed pressure. In: Precision Agriculture’05. 5th European value decreased with increase in time. For higher temperature, there Conference on Precision Agriculture, Uppsala, Swede, Wageningen is a lower drying time, since the rate of evaporation of the product Academic Publishers, 169–174. increases with the temperature. It was also found that in the drying Sacilik, K., Elicin, A. K. (2006). The thin layer drying characteristics of organic process of apple slices that the colour share value dropped signifi - apple slices. Journal of Food Engineering, 73: 281–289. cantly within 30 min, but the moisture content decreased gradually Seiiedlou, S. A.  D. E.  G. H., Ghasemzadeh, H. R., Hamdami, N., Talati, F. for all the varieties in all temperatures. Good coefficient of determi - A. R. A. M. A. R. Z., Moghaddam, M. (2010). Convective drying of apple: nation was established with significant linear relationship at lower mathematical modeling and determination of some quality parameters. International Journal of Agriculture and Biology, 12: 171–178. temperature values of 60°C in colour share values. With increase in Shahin M. A., Tollner E. W., McClendon R. W., Arabnia H. R. (2002). Apple temperature, decrease in the coefficient of determination was noted. classification based on surface bruises using image processing and neural The adjusted coefficient of determination was found to be around networks. Transactions of the ASAE, 45: 1619–1627. 69 per cent for 60°C, 54 per cent for 70°C, and 45 per cent for 80°C Sturm, B., Hofacker C. W., Hensel O. (2012). Optimizing the drying parame- temperatures, respectively, for the linear model of moisture content ters for hot-air–dried apples. Drying Technology-An International Journal, and the colour share value, as more than 50 per cent of change in 30-14: 1570–1582. moisture content was validated by average colour share, which is Vadivambal, R., Jayas, D. S. (2007). Changes in quality of microwave-treated reasonable for natural products. Linear relationship is established as agricultural products-a review. Biosystems Engineering, 98: 1–16. the best fit to identify the change in moisture content through aver - Velic, D., Planinic, M., Tomas, S., Bilic, M. (2004). Influence of airflow veloc - age colour share values. For further research and better results, the ity on kinetics of convection apple drying. Journal of Food Engineering, 64: 97–102. number of samples can be increased and digital information can be Veraverbeke E. A., Verboven P., Van Oostveldt P., Nicolaï B. M. (2003) captured at controlled ambient conditions. Prediction of moisture loss across the cuticle of apple [Malus sylves- trissubsp. mitis (Wallr.)] during storage: part 2: model simulations References and practical applications. Postharvest Biology and Technology, 30: 75–88. Bora, C. G., Lin, D., Bhattacharya, P., Bali, K. S., Pathak, R. (2015). Application Vesali, F., Gharibkhani, M., Komarizadeh, M. H. (2011). An approach to esti- of bio-image analysis for classification of different ripening stages of mate moisture content of apple with image processing method. Australian banana. Journal of Agricultural Science, 7-2: 1916–9752. Journal of Crop Science, 5: 111–115 Contreras, C., Martin-Esparza, M. E., Chiralt, A., Martinez-Navarrete, N. Vibhute, A., Bodhe, S. K. (2012). Application of image processing in agri- (2008). Influence of microwave application on convective drying. Effects culture: a survey. International Journal of Computer Application, 52: on drying kinetics, and optical and mechanical properties of apple and 34–40 strawberry. Journal of Food Engineering, 88: 55–64. Wilhelm, L. R., Suter, D. A., Brusewitz, G. H. (2004). Chapter 10: Drying and Du, C. J., Sun, D. W. (2004). Recent developments in the applications of image dehydration. In: Food and Process Engineering Technology. ASAE, ST. processing techniques for food quality evaluation. Trends in Food Science Joseph, Michigan, pp. 259–284. & Technology, 15: 230–249. Downloaded from https://academic.oup.com/fqs/article-abstract/2/2/105/4995532 by Ed 'DeepDyve' Gillespie user on 26 June 2018

Journal

Food Quality and SafetyOxford University Press

Published: May 12, 2018

There are no references for this article.