Research on the Cable-to-Terminal Connection Recognition Based on the YOLOv8-Pose Estimation Model
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsI have reviewed the manuscript “Research on the Cable-to-Terminal Connection Recognition Based on Pose Estimation”. My comments are as follows:
1. The title should be revised to reflect the explicit use of YOLOv8 pose estimation model in the author’s approach.
2. Line 26, “terminal-to-cable” should read “cable-to-terminal” as the terminal points are usually fixed to the electrical panel.
3. Figure 1 image resolution should be improved as the labels “BM1” and “BM2” may be difficult to decipher for the reader.
4. The introduction section, although well structured should include the main contributions of the study to highlight the unique features of the research.
5. Line 134, “Guibas et al.” cannot be found in the reference section. Please verify alignment of the manuscript with the reference section.
6. Line 202, Equation (5) variables should be well defined. Lambda, theta, and so on.
7. Line 244, Equation (10), variable “v” definition is missing in text.
8. Line 250, suggest section 3 heading change to “Wiring Recognition Methods”
9. In section 4, the authors should state the number of sample images used in the training, validation and test even though a ratio is mentioned. What is the sample size?
10. Line 418, Equation (15). Is Lambda the same variable as used in Eq. (5)?
11. Line 446, the authors should state the actual figure and label the sub graphs accordingly, Fig8. (a), (b), (c), …
12. Line 449, even though the authors draw a general conclusion to the optimal value of “key-point loss weight” in the range 15 to 25, it is not very evident from Figure 8. Perhaps a different graph type may be easier to interpret the overall optimal value range.
13. Comparison with other image recognition approaches, such as Mask R-CNN, HR-Net and Alpha pose is lacking. The authors should include this to show a comparative assessment with the other techniques found in the literature. This will provide context to the final results obtained.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Comments and Suggestions for AuthorsThe article deals with the important issue of real-time monitoring of cable connections to terminals in power transmission substations. The article proposes to use machine learning methods for an automated terminal connection recognition system. The article uses the modern YOLOv8-pose toolkit, which supports 3D recognition of connections, to solve the issue of controlling cable connections to terminals. Using the similarity between human posture position recognition and terminal-cable connection, the paper proposes a position estimation method to recognize terminal-cable connections.
In the introduction, the authors explained in detail the problem that is solved in the article. Section 2 explains the idea behind the YOLOv8-pose model. Section 3 describes the idea of the methods used in the article. Chapter 4 presents the research results and their analysis. The conclusions present a summary of the conducted research and define the prospects for further research.
The results of the research presented in the article under different viewing angles demonstrate the effectiveness of using the YOLOv8-pose model for monitoring reliable cable connection. The article shows the effectiveness of applying the YOLOv8n model to solve the problem of displacement between terminals and cables, which increases the accuracy and reliability of recognition. Having conducted a study on selecting the value of the weight to obtain the best recognition results, the optimal value equal to 22 was determined in the article. The results presented in the article are important for automating the processes of monitoring the reliable operation of electrical substations.
Remark:
Formula (5) (line 202) uses parameters that need to be explained. Formula (15) (line 418) completely repeats formula (5) and already has an explanation of the parameters. It is not necessary to repeat the entry of the same formula. This inconsistency needs to be corrected.
It is necessary to adhere to the same notation of symbols, because in lines 359,364,375 we see the notation mAP50-95, although in formula (14) in line 376 we see 𝑚𝐴𝑃50−95
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors have attended to all the review comments made previously. However, the figures 8,9,10 resolution has deteriorated for some reason. Please also check the figure numbering as there are two fig.9.
Author Response
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.
Comments 1: The authors have attended to all the review comments made previously. However, the figures 8,9,10 resolution has deteriorated for some reason. Please also check the figure numbering as there are two fig.9.
Response 1: Thanks for the review’s comment, we redrawed figures 8, 9, and 10 with high resolution in this resubmssion. In the last manuscript, the second figure 9 should have been labeled as figure 10, we have corrected this typo.
