ORD-YOLO: A Ripeness Recognition Method for Citrus Fruits in Complex Environments

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper proposes a method to estimate citrus fruit grape maturity using an improved algorithm. Here are my considerations to the authors.

Introduction

A section on state-of-the-art research progress on lightweighting of model parameters  for YOLO model needs to be organized in the introduction.

Materiel and methods

Line 119, in the period studied (June 2023 to October 2024) how many harvests did you have?
Line 120 Why was greenhouse citrus cultivation chosen in your study?
Line 192 Why did the authors not use the latest version of YOLO? When this review is being written, the latest version of YOLO is the 9 version. Please explicitly state the advantages of YOLOv9 if compared to YOLOv10 and 11.

Conclusions

Line 595 "The proposed improved citrus fruit maturity detection model, ORD-YOLO, demonstrates outstanding performance in complex orchard environments." Though your study was based in greenhouse?

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This paper proposed a ripeness recognition method for citrus fruits in complex environments. The topic is practically relevant, and the manuscript is generally well-structured with clear problem formulation, methodological innovation, and extensive experimental validation on real-world data. However, there remains room to enhance the work's depth, critical insight, and structural synthesis. The specific comments are as follows:

 

For the research significance and experimental object:

1. The manuscript focuses on the citrus fruits. It is suggested that the key differences in characteristics between the citrus fruits and other fruits be mentioned.
2. It's suggested that the author further clarify the necessity and scientific value of whether the method described has the universality to be extended to other related ripeness recognition research.
3. Please further clarify the practical significance and application value of ripeness recognition for this fruit in the introduction.
4. Please specify in the introduction what the complex environment referred to in the article title is and why these environments have an impact on ripeness recognition.

 

For the experimental validity:

1. It is recommended that the recently proposed methods be compared and verified in the comparative experiments.
2. To better assess the algorithm's robustness and accuracy, additional experiments under natural environmental variations, such as Devices with different resolution, are recommended.

 

For the writing improvement:

1. In line 79-81, the problems summarized for these methods are too broad. It is suggested to refine the existing problems.
2. Please sort out your motivation for each step in detail in Section 3.2. The existing description is too brief.
3. Please elaborate on the parameters in equation 1 again. The current description is not clear enough, for example, the dimension of .
4. It is suggested that the feasibility of the proposed model in practical application should be supplemented in the Discussion part.
5. At present, there are too many words in the conclusion. It is suggested to reduce the text of the conclusion.
6. The text annotations in Figure 9 are too small to be easily legible. It is recommended to enlarge the font size and, if necessary, reduce the number of subfigures to highlight key regions more effectively.
7. Please summarize the main contribution of this article in the introduction section.
8. Please write the arrangement of subsequent chapters in the introduction section of this article.

For the paper structure:

1. To improve the clarity and logical structure of the manuscript, it is suggested to introduce a separate Related Work section. The existing content in the Introduction that pertains to prior research can be moved to this section, allowing the Introduction to focus on motivation and contributions.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors This article proposes an improved YOLOv8 based DL for detecting citrus fruit maturity in complex environments. The model integrates several modules to enhance robustness, accuracy, and real-time inference. The paper is well-structured and addresses a timely challenge in smart agriculture, particularly in automating fruit harvesting. There are some minor revisions that need to be done:

1. Some figures such as: No. 2, 3, 4, 6, 8, 9 have very small fonts or low resolution. Please improve image clarity and increase label sizes for better readability.
2. While the study focuses on the Xingjin Honey Tangerine variety, the model’s generalization to other citrus types is not tested. Consider mentioning future testing on more varieties.
3. In Table 3 a little explanation of training time per epoch or the total training duration may enhance the feasibility.
4. The paper compares ORD-YOLO with YOLO variants and traditional models, but mentioning newer transformer-based detectors such as DETR would improve completeness.
5. A light proofreading pass is recommended.
6. Any actual deployment or edge test was performed, or if this is only theoretical?
7. The paper would benefit from citing very recent work (2024–2025) on lightweight detection models in agricultural contexts.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The author's revisions have responded well to the reviewers' comments, and I have no further suggestions.