Optimizing Deep Learning Algorithms for Effective Chicken Tracking through Image Processing
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors· The work presented in the Manuscript, entitled „Optimizing Deep Learning Algorithms for Effective Chicken Tracking through Image Processing„. There are shortcomings and modifications that should be included in order to enhance the final manuscript for the readers.
· The sentence from line 22 to line 23 should be added in line 26. The results should be added together.
· In line 23 (a loss of less than 0.1). Please modified them, it is not clear.
· Please add the digital results in abstract for the three scenarios such as the performance of them?.
· In introduction section, the novelty of this study should be presented in more details. The introduction is very brief.
· In Materials and Methods, the authors do not clarify where the scientific research took place or what country it was?.
· In line 85. The name of the manufacturer, country and version of POWER® camera should be written.
· In line 104. The goal should be added in introduction.
· More information data about the models like equations and statistical indicators which used to test the performance of the training and validation models should be added in M&M.
· Abbreviations in figure3 must be written in detail in the title of the figure.· Line 192 (3.1. Yolo detection model). The authors should be presented the results of figure 3 with more details.
· Please, write the practical applications of your work in a separate section, before the conclusions and provide your good perspectives.
· Conclusion section is is very brief. Please modified it?
· Please write about the limitations of this work in details in conclusion section.
Comments on the Quality of English LanguageMinor editing of English language required
Author Response
Dear Reviewer,
On behalf of the authors, I would like to thank you for taking the time to review our manuscript. We have made every effort to comply with all your suggestions. Below, we provide detailed responses to each of your notes.
Yours sincerely.
- The sentence from line 22 to line 23 should be added in line 26. The results should be added together.
Answer: Done.
- In line 23 (a loss of less than 0.1). Please modified them, it is not clear.
Answer: We have standardized the presentation of accuracy and loss values.
- Please add the digital results in abstract for the three scenarios such as the performance of them?.
Answer: We added a qualitative description of the results obtained in the three scenarios evaluated.
- In introduction section, the novelty of this study should be presented in more details. The introduction is very brief.
Answer: We have improved the introduction.
- In Materials and Methods, the authors do not clarify where the scientific research took place or what country it was?.
Answer: We have included the location of the experiment in the study.
- In line 85. The name of the manufacturer, country and version of POWER® camera should be written.
Answer: This camera is generic and has a very low acquisition cost (less than 30 dollars), with resellers branding it. We lack information on its manufacturing origin. Our research aimed to utilize low-cost cameras to make chicken monitoring technology economically feasible for the economic realities of Brazilian poultry farms. Unfortunately, the description we included in the article is the best we could manage.
- In line 104. The goal should be added in introduction.
Answer: We rewrote the objectives that were not clear.
- More information data about the models like equations and statistical indicators which used to test the performance of the training and validation models should be added in M&M.
Answer: We have added descriptions of the performance metrics used to evaluate the Yolov8 model.
- Abbreviations in figure3 must be written in detail in the title of the figure.
Answer: We added this information in the figure title.
- Line 192 (3.1. Yolo detection model). The authors should be presented the results of figure 3 with more details.
Answer: We added an analysis of the metrics presented in Figure 3. Additionally, we included in the Materials and Methods chapter a list and description of the set of performance evaluation metrics for the Yolov8 model.
- Please, write the practical applications of your work in a separate section, before the conclusions and provide your good perspectives.
Answer: We took the reviewer's suggestion and created a section for Practical Applications and Perspectives.
- Conclusion section is is very brief. Please modified it?
Answer: We rewrite our conclusions.
- Please write about the limitations of this work in details in conclusion section.
Answer: We add the limitations of this study in the conclusions section.
Reviewer 2 Report
Comments and Suggestions for Authors1. Abstract. Line 23. This results in a precise measure of the chickens' size and shape, with the YOLO model achieving an accuracy above 98% and a loss of less than 0.1. Be consistent. 0.98 and 0.1 or 98% and 10%.
2. Line 59. It is also interesting to highlight the difference between the separate and joint trackers. Mention the advantages and disadvantages of each method.
3. Line 177. How did the authors determine the 60 square pixels as a threshold for area image filtering? This threshold value is situational depending on the size of samples (physical properties of chickens). How did the authors adapt the algorithm for the next samples?
4. Please add in the materials and methods section: Performance Evaluation. Please evaluate the result of object detection completely using the following parameters: precision, recall, and F1-score. Add these parameters in the result section.
5. The detailed analysis of these results reveals significant improvements in the model's performance, achieving an accuracy above 98% and a loss of less than 0.1. Be consistent. 0.98 and 0.1 or 98% and 10%.
6. In the model evaluation in the 3.2 section, the authors should mention quantitatively the prediction result of the object detection in terms of confusion matrices and show the precision, etc. It is also important to calculate AP50% based on the matrices. The authors can compare the results with previous studies in an apple-to-apple comparison.
Author Response
Dear Reviewer,
On behalf of the authors, I would like to thank you for taking the time to review our manuscript. We have made every effort to comply with all your suggestions. Below, we provide detailed responses to each of your notes.
Yours sincerely.
- Abstract. Line 23. This results in a precise measure of the chickens' size and shape, with the YOLO model achieving an accuracy above 98% and a loss of less than 0.1. Be consistent. 0.98 and 0.1 or 98% and 10%.
Answer: Done
- Line 59. It is also interesting to highlight the difference between the separate and joint trackers. Mention the advantages and disadvantages of each method.
Answer: We write the advantages and disadvantages of each method.
- Line 177. How did the authors determine the 60 square pixels as a threshold for area image filtering? This threshold value is situational depending on the size of samples (physical properties of chickens). How did the authors adapt the algorithm for the next samples?
Answer: We have added the following explanatory text to the article.
“The threshold of 60 square pixels for area image filtering was established through a methodical process of trial and error. This value, albeit small, was found to be optimal for our specific context. It is important to note that the size of the chickens, which is the primary subject of our study, is significantly larger than this threshold, with an average area of 19,573 pixels.
Given this substantial difference in scale, the threshold of 60 square pixels is unlikely to introduce any significant distortions or inaccuracies in our analysis. Furthermore, this threshold does not necessitate any adaptations when applied to different samples, as its value is relatively insignificant compared to the size of the chickens.
In essence, the choice of this threshold is a balance between eliminating noise and preserving the integrity of the sample’s physical properties. It is a testament to the robustness of our algorithm that it can accommodate such variations without the need for individual adjustments. We believe this approach enhances the generalizability and applicability of our method across different samples.”
We hope this provides a comprehensive response to your query and we appreciate your insightful review which helps us improve the quality of our work. Please feel free to reach out if you have any further questions or require additional clarification.
- Please add in the materials and methods section: Performance Evaluation. Please evaluate the result of object detection completely using the following parameters: precision, recall, and F1-score. Add these parameters in the result section.
Answer: In the Materials and Methods chapter, we included a list and description of the set of performance evaluation metrics for the Yolov8 model. To evaluate the modified Yolov8, accuracy, precision, specificity, sensitivity, error rate, Kappa coefficient, and F1-score were calculated.
- The detailed analysis of these results reveals significant improvements in the model's performance, achieving an accuracy above 98% and a loss of less than 0.1. Be consistent. 0.98 and 0.1 or 98% and 10%.
Answer: Adjusted.
- In the model evaluation in the 3.2 section, the authors should mention quantitatively the prediction result of the object detection in terms of confusion matrices and show the precision, etc. It is also important to calculate AP50% based on the matrices. The authors can compare the results with previous studies in an apple-to-apple comparison.
Answer: We added Table 1, which contains the performance evaluation of the modified model.
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript was improved according my comments. It can be accepted for publication.
Reviewer 2 Report
Comments and Suggestions for AuthorsThe revised manuscript is now acceptable for publication.
