Fuzzy Classification of the Maturity of the Orange (Citrus × sinensis) Using the Citrus Color Index (CCI)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

You have three equations in the manuscript. In the left-hand side of each equation you have a subscript i, however the right-hand sides do not contain  index i. So what is the definition of i? In lines 368 and 370 you use capital i. Is it true that I and i denote the same index?

Figure 3 is a very good picture. The membership functions are chosen to be linear triangular in Figure 10. Is it possible to use different kind of membership functions in the Takagi-Sugeno inference scheme?

Author Response

Response to the Reviewer Comments

Reviewer 1

1.- You have three equations in the manuscript. In the left-hand side of each equation you have a subscript i, however the right-hand sides do not contain  index i. So what is the definition of i? In lines 368 and 370 you use capital i. Is it true that I and i denote the same index?

R= We are grateful for your observation and have made the necessary adjustments to the equations to emphasize the significance of the subscript "i". To provide further clarification, we have included the following information:

"It should be noted that the subscript 'i' indicates the rule number utilized to calculate the weighted average of the maturity, degree Brix, and firmness using the assigned weights and output levels."

2.- Figure 3 is a very good picture. The membership functions are chosen to be linear triangular in Figure 10. Is it possible to use different kind of membership functions in the Takagi-Sugeno inference scheme?

R= We appreciate your observation. In Figure 10, the membership functions were modified to a Gaussian type. If it is possible to use membership functions of the triangular type, this answer is affirmed by the results obtained from the use of Matlab's Fuzzy Logic and Neuro-Fuzzy Designer Apps. In these results, it was observed that the fuzzy inference model with Gaussian functions for maturity, degrees Brix, and firmness presented the lowest mean square errors, with values of 0.100, 0.289, and 0.250, respectively. The coefficients of determination (R²) were greater than or equal to 0.97. The lowest mean square errors were observed for degrees Brix and firmness, with values of 0.100, 0.289, and 0.250, respectively. Additionally, the coefficients of determination (R²) were found to be greater than or equal to 0.97, indicating a high degree of correlation between the predicted and actual values. Table 5 presents a comparison of the membership functions of the triangular, trapezoidal, bell, and Gaussian types, employing the Takagi-Sugeno inference method illustrated in Figure 10. Moreover, the fuzzy models were analyzed using the color descriptor of the five CCI sub-regions, which corresponded to the dimensions of 3x3, 5x5, 11x11, 21x21, and 31x31.

The authors appreciate the comments and guidance of the reviewer to improve the manuscript. Thank you very much.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors proposed a system for evaluating the ripeness of oranges using computer vision and fuzzy inference system. The proposed methodology was composed of acquisition of RGB images, noise removal, determination of the fruit's centroid, definition of five regions, survey of Citrus Color Index (CCI) by varying pixels and creation of a model to estimate maturity, Brix and firmness.

The article presented a case study, using a sample of 47 Valencia oranges classified into 4 groups of different degrees of maturity. It is presented in detail and clearly, illustrated with figures and tables. Comparisons of one of their models with the best results, with four others found in the literature are presented. The results show that their proposal presented satisfactory results.

I have some considerations.

  -  The results are based on an example. I believe that to validate the proposal, more diverse tests are needed.

-        Lines 59 to 64, 145: References 19, 21, 22, 23 and 53 were not mentioned in the text of the article.

-       Tables 3 and 4, respectively brix grade and firmness of orange are the same. It needs to be justified why they are the same.

-       Table 9: "(Pi et al. 2020) [53]" -  Wouldn't it be "(Li et al. 2020) [53]"?

-       References 2.: It is desirable that "at al" be replaced with the names of other authors.

Author Response

Response to the Reviewer Comments

Reviewer 2

The authors proposed a system for evaluating the ripeness of oranges using computer vision and fuzzy inference system. The proposed methodology was composed of acquisition of RGB images, noise removal, determination of the fruit's centroid, definition of five regions, survey of Citrus Color Index (CCI) by varying pixels and creation of a model to estimate maturity, Brix and firmness.

The article presented a case study, using a sample of 47 Valencia oranges classified into 4 groups of different degrees of maturity. It is presented in detail and clearly, illustrated with figures and tables. Comparisons of one of their models with the best results, with four others found in the literature are presented. The results show that their proposal presented satisfactory results.

I have some considerations.

1.- The results are based on an example. I believe that to validate the proposal, more diverse tests are needed.

R= We appreciate your observation to improve the validation method of the results obtained. The number of samples of Valencia oranges was increased from 47 to 75 based on similar research such as Domingues [50], Olmos [52], Habibi [44], and Villaseñor [36]: According to their results, it was possible to determine physicochemical variables such as degrees Brix on fruits using a quantity of samples equal to or less than that reported in this work. These references are included in the paper.

[50] Domingues, A.R.; Marcolini, C.D.M.; Gonçalves, C.H.d.S.; Gonçalves, L.S.A.; Roberto, S.R.; Carlos, E.F. Fruit Ripening Development of  Valencia Orange Trees Grafted on Different ‘Trifoliata’ Hybrid Rootstocks. Horticulturae 2021, 7, 3. https://doi.org/10.3390/horticulturae7010003.

[52] Olmo, M.; Nadas, A.; García, J.M. Nondestructive Methods to Evaluate Maturity Level of Oranges. Sens. Nutr. Qual. Food Nondestruct. 2000, 65, 365–369.

[44] Habibi, F.; Guillén, F.; Serrano, M.; Valero, D. Physicochemical Changes, Peel Colour, and Juice Attributes of Blood Orange Cultivars Stored at Different Temperatures. Horticulturae 2021, 7, 320. https://doi.org/10.3390/horticulturae7090320.

[36]. Villaseñor-Aguilar, M.-J.; Bravo-Sánchez, M.-G.; Padilla-Medina, J.-A.; Vázquez-Vera, J.L.; Guevara-González, R.-G.; García-Rodríguez, F.-J.; Barranco-Gutiérrez, A.-I. A Maturity Estimation of Bell Pepper (Capsicum annuum L.) by Artificial Vision System for Quality Control. Appl. Sci. 2020, 10, 5097. https://doi.org/10.3390/app10155097

 

2.- Lines 59 to 64, 145: References 19, 21, 22, 23, and 53 were not mentioned in the text of the article.

R= Thank you for your observation, The references 19, 21, 22, 23, and 53 have already been mentioned in the text of the article.

3.- Tables 3 and 4, respectively brix grade and firmness of orange are the same. It needs to be justified why they are the same.

R= Thank you for your comment, Tables 3 and 4 have been changed. The mean squared error (MSE) columns have been corrected.

4.- Table 9: "(Pi et al. 2020) [53]" -  Wouldn't it be "(Li et al. 2020) [53]"?

R= We apologize for the mistake, Table 9 has been changed in this new version of the paper by correcting the reference from Pi et al. 2020 to Li et al. 2020.

5.- References 2.: It is desirable that "at al" be replaced with the names of other authors.

R= Thank you for the comment, reference 2 has been corrected. The "et al" has been changed to the names Freitas, E.d.; Novello, D.

 

 

The authors appreciate the comments and guidance of the reviewer to improve the manuscript. Thank you very much.

 

Author Response File: Author Response.pdf