with the collaboration of Iranian Food Science and Technology Association (IFSTA)

Document Type : Full Research Paper

Authors

1 Department of Mechanics of Biosystems Engineering, Faculty of Agricultural Engineering and Rural Development, Agricultural Sciences and Natural Resources University of Khuzestan, Ahvaz, Khuzestan Iran.

2 Department of Mechanics of Biosystems Engineering, Faculty of Agriculture, Tarbiat Modares University, Tehran, Iran.

Abstract

Introduction: The development of brown spots on banana peel has a notable effect on the texture, color and taste of this fruit. So that the appearance of these spots reduces the quality of the fruit and affect its sale market. In recent years, in order to evaluate the quality and classification of agricultural products, the various systems based on computer vision technology have been widely considered. These systems as the computer image analysis methods have been successful in measuring the visual quality of different products (Riyadi et al., 2007; Roseleena et al., 2011; Rodriguez-pulido et al. 2012). According to research by Probha and Kumar (2015), the extracted color properties from the banana image were more effective than other features in identifying the different stages of the banana ripening. Also, Mendoza and Aguilera (2004) detected the different stages of banana ripening based on the color, texture parameters and the distribution of brown spots on banana peel using image processing technique with a precision of 98%. Nadafzadeh et al. (2018) designed a non-linear mathematical model using the Genetic Programming (GP) to predicting and evaluating the activity of polyphenol oxidase enzymes (PPO) and peroxides (POD) during the browning process of the banana peel; using the extracted parameters from image as inputs of proposed model, the correlation coefficients to predicting of PPO and POD enzymes were obtained 0.98 and 0.97, respectively.
The aim of this study was to investigate the changes of color, dimensions and chemical parameters of several banana fruit groups (different in terms of appearance) as well as their marketability (the total acceptance of fruit) by Gaussian regression model (GPR) during the storage period. Therefore, using the proposed method in this research, the required product can be available according to the consumer demand.
 
Materials and Methods: In this study, one hundred banana samples were prepared from a market on the first day of the experiments. Samples were different in terms of shape and size, and were classified into 5 different groups. Group A had small size and curvature; B group compared to Group A had more curvature; the curvature of the samples in the group C was high, and in terms of size were medium. While the size of the bananas in group D was large, they had a small curvature. Also, the features of the group E were similar to the group D, but the curvature was greater in this group (group E). All of the samples were kept at the ambient temperature (25° C) away from the direct light for 7 days. During the days of experiments (days 0, 2, 4 and 6), five samples were examined from each group: after taking images of samples under the constant light conditions, and performing of manual measurements, they were subjected to destructive tests (laboratory tests) and sensory tests. After the images acquisition of samples, the preprocessing operations such as image enhancement, noise removal by the area opening, and the implementation of the image segmentation process using the method of Otsu adaptive thresholding were conducted (Gonzalez et al., 2004). Finally, 11 color parameters (R, G, B, L, a, b, h, s, v, C, H) and 4 dimensional characteristics (diameter, curvature radius, long and small length) were extracted from each image. In the laboratory method, the TSS value was measured by a digital refractometer, and amount of pH and acidity were also measured by a fruit juice analysis titrator. Eventually, in order to investigate the changes of measured parameters, statistical analysis was performed in a randomized complete block design by SAS 9.3 software at a significance level of 5% using Duncan's multiple comparison test.
 
Results and discussion: Gradually along with the appearance of dark spots on the banana peel, many of the qualitative parameters such as the color, dimensions and chemical features were changed during the storage period. According to results of the Duncan's multiple range test, the values of color coordinates R, G, B, L, b, h, v, C, and H gradually reduced, and the values of these parameters were significant in all the experiments days (p<0.05). The parameter S also had a decreasing trend during the storage period, and the changes of this parameter was significant in the first days of the experiments compared to the ending days; during this period, the color parameter a increased significantly. Due to the changes of the banana fruit texture, the amount of the curvature radius, the small and large lengths, total soluble solids, pH and total titration acidity gradually decreased. Based on the results of the statistical analysis, there were no significant differences between dimensional parameters measured by non-destructive method and manual measurement (p>0.05). It is worth noting that in this study, the spent time to conduct the manual measurements of the dimensional parameters of a banana sample was 510 seconds, while all of these measurements were performed using a digital image processing method at 1.015 seconds. Therefore, it can be said that when the number of samples is high, using of the proposed method is also very cost-effective in terms of time, and it has high accuracy during the measurement. In the sensory evaluation, the results show that the best and most acceptable group of bananas were groups C, D and E, which had long size and low curvature; these groups of bananas had delicious texture, desirable flavor and low levels of brown spots on their peel. In the following, the non-destructive parameters were used to the development of Gaussian regression model (GPR), and finally, it was shown that the quality of banana fruit as well as its marketability (the total acceptance of fruit) are predictable during the storage period by GPR with a correlation coefficient of 0.91, MAPE (20.47), RMSE (0.43), SRE (0.71) and RAV (0.20).
The appearance quality of the banana fruit is very effective in its acceptability for customer. In this research, the image processing technique as a non-destructive method was used to extract a set of color (R, G, B, L, a, b, h, s, v, C and H) and morphological properties (diameter, curvature radius, long length and small length) from banana image in order to evaluate its quality during storage. According to the results of Duncan's statistical analysis at the probability level of 5% and Pearson correlation results, the most suitable parameters were chosen to apply in Gaussian regression model. The results showed that the image processing technique is capable to evaluating the changes of color and dimensional parameters of banana fruit, and also the proposed model have a satisfactory performance (R2=0.91) in predicting the overall acceptance parameter of the banana.
 

Keywords

سلطانی کاظمی، م.، آبدانان مهدی‌زاده، س. ، نداف‌زاده، م. (1396). برآورد میزان دو آنزیم‏ PPO و POD موز با استفاده از پردازش تصاویر دیجیتالی و آنالیز رگرسیونی چندگانه در طول دوره انبارمانی. فناوری‌های نوین غذایی، 5 (4)، 597-612.
نداف زاده، م.، و آبدانان مهدی‌زاده، س. (1396). تعیین مناسب ترین فضای رنگی به منظور تعیین هوشمند تنش آبی درگیاهان گلخانه‌ای (مطالعه موردی: حُسنِ‌یوسف). مهندسی بیوسیستم ایران، 48 (4)، 418-407.
نعمتی‌نیا، ا.، آبدانان مهدی‌زاده، س.، و ناصحی، ب. (1396). اندازه گیری پارامترهای رنگ در اسپاگتی با استفاده از سیستم بینایی ماشین. مجله علوم و صنایع غذایی ایران، 14 (73)، 81-71.
Cho, J.S., Lee, H.J., Park, J.H., Sung, J.H., Choi, J.Y., Moon, K.D., 2016. Image analysis to evaluate the browning degree of banana (Musa spp.) peel. Food Chem. 194, 1028–1033.
CK Narayana, MM Mustaffa. (2006). Influence of maturity on shelf life and quality changes in banana during storage under ambient condition. Indian Journal.
D. Surya Prabha , J. Satheesh Kumar(2015) Department of Computer Applications, School of Computer Science and Engineering, Bharathiar University, Coimbatore 641 046, Tamil Nadu, India
Garcia-Mateos, G., Hernandez-Hernandez, J. L., Escarabajal-Henarejos, D., Jaen-Terrones, S. & Molina-Martinez, J. M.)2015(. Study and comparison of color models for automatic image analysis in irrigation management applications. Agricultural Water Management. 151, 158-66.
Gonzalez, R.C., Woods, R.E., 2009. Digital Image Processing using MATLAB, 3th ed., Pearson Education India, pp. 1-954.
Gonzalez, R.C., Woods, R.E., Eddins, S.L., 2004. Digital Image Processing Using MATLAB. Pearson Education, India.
Goudarzi, M., Madadlou, A., Mousavi, M., Emam-Djomeh, Z. (2014). Formulation of apple juice beverages containing whey protein isolate or whey protein hydrolysate based on sensory and physicochemical analysis. Society of dairy technology., 67, 1-9.
He, H., & Siu, W. C. (2011, June). Single image super-resolution using Gaussian process regression. In Computer Vision and Pattern Recognition (CVPR), 2011 IEEE Conference on (pp. 449-456). IEEE.
Jaiswal, P., Jha, S.N., Kaur, P.P., Bhardwaj, R., Singh, A.K. and Wadhawan, V., 2014. Prediction of textural attributes using color values of banana (Musa sapientum) during ripening. Journal of food science and technology, 51(6), pp.1179-1184.
Mendoza F, Aguilera JM (2004) Application of image analysis for classification of ripening bananas. J Food Sci 69:E471–E477
Nadafzadeh, M., Mehdizadeh, S. A., & Soltanikazemi, M. (2018). Development of computer vision system to predict peroxidase and polyphenol oxidase enzymes to evaluate the process of banana peel browning using genetic programming modeling. Scientia Horticulturae, 231, 201-209.
P Rajkumar, N Wang, G Elmasry, GSV Raghavan, Y Gariepy. (2012). Studies on banana fruit quality and maturity stages using hyperspectral imaging. Journal of food engineering 108 (1), 194-200.
Pilar Cano, M Antonia Martin, Carmen Fuster. (1990). Effects of some thermal treatments on polyphenoloxidase and peroxidase activities of banana. Journal of the science of food and agriculture 51 (2), 223-231.
Quevedo R, Mendoza F, Aguilera JM, Chanona J, Gutierrez-Lopez G (2008) Determination of senescent spotting in banana (Musa cavendish) using fractal texture fourier image. J Food Eng 84:509–515.
Quevedo, R., Diaz, O., Ronceros, B., Pedreschi, F., Aguilera, J.M., 2009. Description of the kinetic enzymatic browning in banana (Musa cavendish) slices using non-uniform color information from digital images. Food Res. Int. 42 (9), 1309–1314.
Riyadi S, Ashrani A, Mohd R, Mustafa M, Hussain A (2007) Shape characteristics analysis for papaya size classification. In: Proceedings of the 5th Student Conference on Research and Development (SCOReD’07), Selangor, Malaysia. IEEE. pp. 371–375.
Rodriguez-Pulido FJ, Gomez-Robledo L, Melgosa M, Gordillo B, Gonzalez-Miret ML, Heredia FJ (2012) Ripeness estimation of grape berries and seeds by image analysis. Comp Elect Agric 82: 128–133.
Roseleena J, Nursuriati J, Ahmed J, Low CY (2011) Assessment of palm oil fresh fruit bunches using photogrammetric grading system. Intern Food Res J 18(3):999–1005.
Sanaeifar A., Bakhshipour A. de la Guardia M. (2016). Prediction of banana quality indices from color features using support vector regression. Talanta, 148, pp.54-61.
Singh HP (2010) Dynamics and Co-kinetics of banana research and Development in India. In: Mustaffa MM (ed) Proceedings of the global conference on meeting the challenges in banana and plantain for emerging biotic and abiotic stress, Trichy, Tamil Nadu, India. ICAR, New Delhi, pp 1–14.
Sochi, T (2016). Tensor Calculus Made Simple. Create Space Independent Publishing Platform. p. 170.
Yoruk, R., Yoruk, S., Balaban, M.O., Marshall, M.R., 2004. Machine vision analysis of antibrowning potency for oxalic acid: a comparative investigation on banana and apple. J. Food Sci. 69 (6), 281–289.
Zomo, S. A., Ismail, S. M., Shah Jahan, M., Kabir, K. and Kabir, M. H., 2014. Chemical Properties and Shelf Life of Banana (Musa sapientum L.) as Influenced by Different Postharvest Treatments. The Agriculturists 12(2): 6-17.
Zulkifli N, Hashima N, Abdan K, Hanafi M. (2019). Application of laser-induced backscattering imaging for predicting and classifying ripening stages of “Berangan” bananas. Computers and Electronics in Agriculture. 160: 100–107.
CAPTCHA Image