ODN-Pro: An Improved Model Based on YOLOv8 for Enhanced Instance Detection in Orchard Point Clouds
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsMinor Issues:
- Extra paragraphs have been created in the introduction section. Paragraphs that cover similar topics or are directly related to each other should be merged.
- In the introduction section, sentences with the same meaning are repeated. The sentence "Constructing a detailed digital map is crucial for guiding autonomous vehicles in smart agriculture." repeats the information provided in the previous sentence. This sentence should be removed.
- In the Data Acquisition section, the term "rosbag" is used but not defined. Its purpose should also be specified.
- The Data Augmentation section does not provide sufficient detail on how each data augmentation technique is applied. For example, specific parameters or implementation details of random point sampling and point cloud transformation techniques are not explained.
- Some figures are not referenced at all in the text (e.g., Figure 8).
- While some figure references are abbreviated (e.g., "Fig.1"), others are not (e.g., Figure 9).
- Figure 7 and Figure 8 are labeled with the same name ("The structure of the ECA block."). If there's a logical explanation for this, it should be included in the text.
Major Issues:
- The section "Related Work" is missing from the text. Studies in the literature that focus on solving the same problem should be described in this section. This would highlight the significance of the study in the literature. Comparisons with these studies would also reveal the performance rate.
- Applying too much data augmentation to the dataset can cause the model to overfit the augmented data, decreasing its ability to generalize to real-world data. The authors have overlooked the problem of overfitting. Techniques like cross-validation should be used to keep this under control.
- The section introducing the main contribution of the paper, ODN-Pro, is not explanatory. The structure shown in Figure 9, which is the abstract form of the paper, needs to be detailed. This section should reveal the unique aspects of the paper. The structural features of the network within the ECA and EVC blocks and how this structure was formulated should be stated with reasons and evidence.
- The outputs of the study have only been compared with YOLOv8. There has been no comparison with studies in this field in the literature. A comparison is needed to measure the success of the method with other studies specified in Related Works.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript is titled "ODN-Pro: An Improved Model Based on YOLOv8 for Enhanced Instance Detection in Orchard Point Clouds." The paper addresses a crucial aspect of unmanned orchard management by proposing an efficient method for fruit tree detection in orchard point cloud maps. However, after careful evaluation, it was identified that the manuscript has several issues that need to be addressed.
The literature review section is currently limited in scope. The authors should conduct some more reviews of relevant studies in the fields of orchard point cloud mapping, deep learning for object detection in agriculture, and related areas (lines 65–68). Add more studies that are related to the present study.
The manuscript lacks clarity in presenting the proposed method. The authors should provide a more structured and cohesive explanation of the methodology.
The methodology section. The authors should provide more details about the figure title below the figure (Figs. 4, 5, and 6) For example, in Fig. 5, briefly describe the functions of the neck, head, and backbone.
In line 93, the authors showed three orchards: a lychee orchard and a citrus orchard, but unfortunately, in the results section or discussion, the reader did not find the explanation for three different areas.
The paper lacks a dedicated discussion section to provide a comprehensive analysis and interpretation of the results. It is crucial to discuss the implications of the findings, compare them with existing literature, and highlight any unexpected outcomes or limitations.
Line 305 Please add the discussion section where authors explain the study results in comparison with other studies. The manuscript lacks a discussion on the implications of the proposed framework, as well as a clear identification of its limitations. The authors should discuss how the implementation of the ODN-Pro model may impact orchard management practices and suggest areas for improvement.
After Lines 336, a section could add on future research directions, which is essential to guide subsequent studies in this domain.
Suggestion
In the abstract, line 11, please delete the words in this paper; the authors already mention these words in line 9.
Line 13: Please change "This paper" to "This study."
In the methodology section, please add a picture of the study sites and show the three orchards.
In the manuscript, authors many times used the words” in this paper,” and we’’ please use these words where they are more needed.
Comments on the Quality of English Languageneed to be improved.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe article titled ODN-Pro: An Improved Model Based on YOLOv8 for Enhanced Instance Detection in Orchard Point Clouds contains new information related to orchard mapping.
It fits into current trends, but it would be worth it for the Authors to indicate whether it is universal and can also be used for other trees, in other conditions, for other objects.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsDear editor and authors:
After evaluating the manuscript “ODN-Pro: An Improved Model Based on YOLOv8 for Enhanced Instance Detection in Orchard Point Clouds”, my decision is: Minor revision.
The introduction is not enough to justify the research. The justification presented in the manuscript does not provide grounds to convince the reader that the research is relevant, especially considering that the manuscript may be published in a journal in the area of agricultural sciences. The problem is not presented clearly, making it difficult to understand the importance and relevance of this study.
The methodology needs further clarification of the proposed procedures. It is clear that some basic information and bibliographic references are missing to ensure greater scientific robustness.
When analyzing the results and discussion of the manuscript, it is not clear whether the models performed well for the cultures tested. In fact, it does not even appear to be a manuscript that is being considered for publication in an agricultural sciences journal. Furthermore, the results were not discussed and justified in the light of science. The authors do not present any weaknesses in the research. What future studies are needed?
Here are other issues that were identified that justify my decision:
Line 22: Keywords should not use terms that have already been included in the title of the work.
Line 24: What is the agricultural area and what tree species are most cultivated in this country? Please use the statistical results to explain.
Line 24: References must be listed in the order they appear in the manuscript. Therefore, it is expected that it should be [1, 2] instead of [39, 19]. This must be corrected throughout the text.
Line 26: The citations are outside the norms, instead of [26], [30], [22] it should be [22, 26, 30]. This must be corrected throughout the text.
Line 30: Which orchards? Give an example with species and conditions.
Line 39: Choose short sentences, as short sentences allow for a better understanding of the text. Think of it this way: (i) Phrase serves to bring ideas about the same subject; (ii) Ideas organized into sentences must close a subject: beginning, middle and end; (iii) So paragraphs serve to separate subjects; (iv) Therefore, paragraphs with one sentence (only an idea of the subject) are not common.
Line 77: Make it clear what your hypotheses and objectives of your study are.
Line 93: Characterize the locations (geographical coordinates), crops (species, phenological stages, size, spacing, management, etc.), soils and climates.
Line 112: Items 2.2., 2.3.1., 2.3.2., 2.3.3., 2.4. and 2.5. present many procedures that are not supported by the literature. The procedures must be in agreement with the literature to show greater robustness of the research carried out.
Line 114: The “paper” is just a document and it does not use and/or do anything. In fact, it was the study or research researchers who decided on the procedures adopted. This must be corrected throughout the text.
Line 130: Fix 2.3.1 “feture”
Line 237: Explain the amount of data for each location (species). Were the data analyzed separately or together? In this study, is the data divided in relation to training, validation and testing of the models? Give a general explanation about the handling of information.
Line 245: What was the processing time for the activities carried out?
Line 247: Change “Results” to “Results and discussion”.
Line 299: Table x?
Given the above, some inconsistencies can be seen and my recommendation is “Minor revision”.
I encourage authors to submit a new version of the manuscript and, if they agree, I would like to receive responses to any considerations they disagree with or do not address!
Good job.
Ad hoc reviewer
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 5 Report
Comments and Suggestions for AuthorsThe work submitted for review contains very important issues regarding precise mapping of plots in perennial crops. The developed method of mapping plots allows you to eliminate inconvenience and inaccurate tracking of plants, which in turn will allow for accurate, impersonal management of crops. There are no substantive comments. He believes that the research methods used are appropriate and justified. The topic discussed is of great practical importance and is very much needed in modern farming conditions.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors have provided acceptable responses to all comments. As a result, the major topics and uncertainties addressed within the paper have been resolved. Lastly, it should be noted that the paragraph under the "2.4. Evaluation Metrics" heading is identified as being sourced from another article through the iThenticate report. Proper citation should be provided for the information taken from this source.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe paper addresses a crucial aspect of unmanned orchard management by proposing an efficient method for fruit tree detection in orchard point cloud maps. However, after careful evaluation, it was identified that the manuscript has several issues that need to be addressed.
1. The literature review section is currently limited in scope. The authors should conduct some more reviews of relevant studies in the fields of orchard point cloud mapping, deep learning for object detection in agriculture, and related areas (lines 65–68). Add more studies that are related to the present study.
2. The manuscript lacks clarity in presenting the proposed method. The authors should provide a more structured and cohesive explanation of the methodology.
3. The methodology section. The authors should provide more details about the figure title below the figure (Figs. 4, 5, and 6) For example, in Fig. 5, briefly describe the functions of the neck, head, and backbone.
4. In line 93, the authors showed three orchards: a lychee orchard and a citrus orchard, but unfortunately, in the results section or discussion, the reader did not find the explanation for three different areas.
5. The paper lacks a dedicated discussion section to provide a comprehensive analysis and interpretation of the results. It is crucial to discuss the implications of the findings, compare them with existing literature, and highlight any unexpected outcomes or limitations.
- Line 305 Please add the discussion section where authors explain the study results in comparison with other studies. The manuscript lacks a discussion on the implications of the proposed framework, as well as a clear identification of its limitations. The authors should discuss how the implementation of the ODN-Pro model may impact orchard management practices and suggest areas for improvement.
7. After Lines 336, a section could add on future research directions, which is essential to guide subsequent studies in this domain.
Suggestion
In the abstract, line 11, please delete the words in this paper; the authors already mention these words in line 9.
Line 13: Please change "This paper" to "This study."
In the methodology section, please add a picture of the study sites and show the three orchards.
In the manuscript, authors many times used the words” in this paper,” and we’’ please use these words where they are more needed.
Comments on the Quality of English LanguageExtensive editing of English language required
Author Response
In this round of reviews, the author found that the reviewers' comments were consistent with those from the initial submission. Nevertheless, the author diligently revised the manuscript in accordance with the reviewers' suggestions and resubmitted the latest revised article. Thanks to the reviewers for their time and suggestions. All feedback and suggestions have been used to improve this research work. Detailed responses and revisions are listed in the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsThe authors have presented an improved version of the manuscript and I understand that it is suitable for publication.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
