Image Segmentation of Cucumber Seedlings Based on Genetic Algorithm

Round 1

Reviewer 1 Report

The manuscript entitled "Image segmentation of cucumber seedlings based on genetic algorithm" has well written and within the scope of journal. The findings are interesting using the optimization techniques. However, there need minor revisions in the manuscript. 

1. The proper justification of the topic at the end of Introduction section.

2. Correct occasionaly grammatical mistakes.

3. Proper formating of the as per template of the journal require.

4. A flowchart methodology of the work done is require for clear understanding.

Author Response

We thank you for the critical comments and helpful suggestions. We have taken all these comments and suggestions into account, and have made major corrections in this revised manuscript. The proposed method has shown excellent performance through experiments, and the paper has been edited in English by MDPI, a formal language polishing organization, to improve the English language and style of my paper. We have revised and improved the article according to your and other reviewers' comments.

Point 1: The proper justification of the topic at the end of Introduction section.

Response 1: ​We have modified and improved this section and supplemented the workflow of the target dosing robot so that the reader can better understand the role of this study in the target spraying task.

Point 2: Correct occasionaly grammatical mistakes.

Response 2: Thank you for the suggestion and the paper has been edited in English by MDPI, a formal language polishing organization, to improve the English language and style of my paper.

Point 3: Proper formating of the as per template of the journal require.

Response 3: We have revised our article according to the template of the journal require.

Point 4: A flowchart methodology of the work done is require for clear understanding.

Response 4: We briefly introduced the flowchart of the algorithm in section 2.1, and introduced the specific implementation of the algorithm in the following sections. And this article has been revised according to the suggestions of the reviewers.

Reviewer 2 Report

Dear authors,

I congratulate you for the article "Image segmentation of cucumber seedlings based on genetic algorithm". It is an interesting article and the method proposed by you must be validated in time.

Author Response

We thank you for the critical comments and helpful suggestions. We have taken all these comments and suggestions into account, and have made major corrections in this revised manuscript. The proposed method has shown excellent performance through experiments, and the paper has been edited in English by MDPI, a formal language polishing organization, to improve the English language and style of my paper. We have revised and improved the article according to your and other reviewers' comments.

Reviewer 3 Report

1. Paper is good.

2.what are the three lighting conditions with percentage of illumination or light ?

3.compare recent published technique.

Author Response

We thank you for the critical comments and helpful suggestions. We have taken all these comments and suggestions into account, and have made major corrections in this revised manuscript. The proposed method has shown excellent performance through experiments, and the paper has been edited in English by MDPI, a formal language polishing organization, to improve the English language and style of my paper. We have revised and improved the article according to your and other reviewers' comments.

Point 1: Paper is good.

Response 1: Thank you for your comments on this article

Point 2: what are the three lighting conditions with percentage of illumination or light ?

Response 2: We do not specifically measure specific proportions and values for the three lighting conditions. However, by referring to other people's research, we selected three time points in May for image acquisition: noon (direct sunlight, no cloud), morning or cloudy day, and dusk. We divided the light intensity of these three time points into strong light, normal light, and weak light. On the one hand, this method enriches the sample pixels. On the other hand, it can also be used to verify the robustness of the proposed algorithms.

Point 3: compare recent published technique

Response 3: By comparing with other recent techniques, our proposed method has higher accuracy and stronger robustness, which can be reflected in the introduction and experimental results.

Reviewer 4 Report

The authors reported a Genetic Algorithm-based segmentation approach for imaged cucumber seedlings. The results prove the accurate applicability of the proposed approach under different lighting conditions, but solely for a single crop species with specific morphometric characteristics. The innovativeness of the proposed algorithm is unclear with respect to other available image-based segmentation techniques. In my opinion, the quality of the manuscript is not sufficient to be published in the present form. However, I encourage the authors to improve their work for future submission of the obtained results. Several comments and suggestions are listed as follows:

1)      The topic investigated by the authors is of interest, but the work is probably outside the focal point of Sustainability journal.

2)      The whole structure of the article is quite confusing. Indeed, the main sections should be Introduction, Materials and methods, Results, Discussion and Conclusions. Likewise, the English language and style need to be extensively edited for a clear understanding of the manuscript.

3)      Line 29: Please, avoid using “etc.” throughout the manuscript.

4)      Into the Introduction section the authors should be better introduce the operation of (e.g.) target dosing robots to better understand how the results of this study could help in their technological advancement in the near future. Moreover, it could be useful to briefly illustrate the pros and cons of the sensors commonly used for image acquisition (e.g., multispectral, hyperspectral and RGB cameras) so as to make the reader understand why an RGB camera was used in this study.

5)      At present, many algorithms are already available for the segmentation of target plant tissues from the background which were successfully applied both in the greenhouse and in the field under controlled and non-controlled light conditions. This aspect should be introduced more in detail by the authors to define the innovation of this work. In this context, all the previous studies mentioned in the Introduction section should be included the relevant results in terms of accuracy, required computation time and generalization to different plant species and growing conditions for a better comparison with the findings of this work.

6)      In the final part of the Introduction section, the authors state that one of the major innovations of the method proposed in the present study is the segmentation speed. However, no data about the required computation time is reported in the following Sections.

7)      Line 46: authors should briefly explain the mechanism of these algorithm for a better comprehension by the reader.

8)      Line 52: please remove the period in the middle of the sentence.

9)      Line 59: please clarify why the cucumber species was chosen for experimentation. Moreover, authors should explain in detail the growing conditions of the tested plants.

10)   Lines 59-60: the sentence “In agricultural production, …” is completely unrelated to the rest of the paragraph.

11)   Line 62: please replace “filming” with “scanning”

12)   Please specify if the camera for the images acquisition was fixed or handheld.

13)   Why did the authors not use a colorimetric reference for the image acquisition? Please specify

14)   Lines 68-70: The intensity of the light inside the greenhouse has not been measured and/or standardized and therefore the lighting conditions are not uniform for the imaged 20 replicas of each thesis.

15)   Line 73: the authors should specify what do they mean for “unified”.

16)   Line 74: please replace “1.2” with “2.2” and maybe “flow” with “workflow”

17)   Lines 97-98: authors should better specify how the different terms were calculated.

18)   Line 114: please specify that this procedure was carried out manually.

19)   Line 127: please insert a valid reference for the formula.

20)   In my opinion, authors should better specify how different coefficients of Equation (6) were computed.

21)   Line 148: please replace (6) with (7).

22)   Line 151: This sentence is not appropriate for Material and Methods.

23)   Lines 247-250: the procedure to reduce the errors of the algorithm reduces its applicability to other cultivars or species with morphological characteristics different from those of the plants taken into analysis.

24)   The Discussion section is missed.

25)   The caption of all the images of the manuscript should be more informative.

26)   Line 329-330: I do not agree with this statement. Standard pixel segmentation must be done manually to calibrate the algorithm.

 

27)   Line 330: please replace “artificial” with “manual”.

Author Response

请看附件。

Author Response File: Author Response.pdf

Round 2

Reviewer 4 Report

The author has made a great effort to revise the article, but some issues need to be considered for publication. In particular, the Discussion section is still missed. Comments are listed as follows:

 

Line 28: I suggest the authors to replace the sentence “such as in weed identification, crop disease detection” with “such as in weed identification and crop disease detection”.

 

Lines 30, 98 and so on: please, avoid using “etc.” and check it throughput the manuscript.  

 

Line 40: please avoid using the “we” personal form and check it throughout the manuscript.

 

Lines 47-48: please, insert a space between the word and the parenthesis. Check it throughout the manuscript.

 

Line 60-61: Authors could merge the two sentences as follows: “Accurate plant identification is fundamental to the robot-based spraying mission, which typically consists of image acquisition, target identification, localization and treatment distribution”.  

 

Section 2.1: I suggest to the authors to modify the order of the first sentences as follows: “In agricultural production, when the cucumber seedling has one or two true leaves, ethephon is sprayed on the leaves multiple times to accelerate female flower differentiation. In this study, cucumber seedlings are taken as the target object. Plants grown in a seedling trial inside the glasshouse of the Guantang Experimental Farm of Zhejiang A&F University. XXX (replace XXX with the total number of imaged plants) seedlings were scanned when the second true leaf appeared using the OV560 camera module from SingTown”. Subsequently, authors should continue with the other sentences related to the shooting system already included in Section 2.1.  

 

Line 111-112: I don't think the word “obvious” is the most appropriate. Please, replace it.

 

Lines 160, 167: Maybe it is better if authors use italics for “a, b, c and d” parameters. Please, uniform it throughout the manuscript.

 

Section 5.1.: This section is not part of the Results. Please, include the text in a separate section titled "Statistical analysis".

 

The Discussion section is still missing. The authors should compare their results with those of previous studies to better determine and underline the effective novelty and utility of the proposed methodology. For example, the authors should include in the Discussion section by how much their algorithm improves the performance of similar algorithms developed previously. Moreover, authors could argue about the applicability of their algorithm to other plant species with a different morphology, colorimetric characteristics and age, as well as the potential application in field conditions. Finally, the authors could speculate on possible future improvements to solve the problems of the proposed methodology or to improve its usability (e.g, by developing a mobile app for infield application).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf