A New Orientation Detection Method for Tilting Insulators Incorporating Angle Regression and Priori Constraints
Round 1
Reviewer 1 Report (Previous Reviewer 1)
Point 1: The authors have addressed most of my concerns. I would suggest the authors adjust the font size or color of Figure 6 and Figure 7 to make it clear.
Author Response
Dear Editors and Reviewer：
Thank you for reviewing and commenting on our paper " A New Bolt defect identification method incorporating attention mechanism and wide residual networks" (Manuscript No. 2073401). Your suggestions are important for the improvement of the study of this paper, and we have carefully studied each comment and made corresponding corrections to the manuscript, and we hope that these efforts will be recognized by you. The detailed responses are listed accordingly below. For convenience, all responses are marked in blue and follow the corresponding comments, and all changes in the manuscript are highlighted in yellow.
We would like to submit the paper for reconsideration for publication in Sensors. Please feel free to let us know if there are any problems or questions about our paper. Thanks again.
- The authors have addressed most of my concerns. I would suggest the authors adjust the font size or color of Figure 6 and Figure 7 to make it clear.
Response：
Thank you for your comments. We apologize for the trouble caused to you. The reason for the lack of clarity in Figure 6 and Figure 7 is that the image resolution was set to 220 dpi, and we re-adjusted the image resolution to 600 dpi to ensure the clarity of Figure 6 and Figure 7.
Thank you again for your recognition of our last revision.
Author Response File: 
Author Response.pdf
Reviewer 2 Report (Previous Reviewer 2)
In general, the revised article has been improved compared with the previous version. Errors such as grammar and symbols in the paper were corrected, and issues raised such as the meaning of each curve and the units of coordinates in Figure 3 were explained and clarified.
Author Response
Dear Editors and Reviewer：
Thank you for reviewing and commenting on our paper " A New Bolt defect identification method incorporating attention mechanism and wide residual networks" (Manuscript No. 2073401). These suggestions are important for the improvement of the study of this paper, and we have carefully studied each comment and made corresponding corrections to the manuscript, and we hope that these efforts will be recognized by you. The detailed responses are listed accordingly below. For convenience, all responses are marked in blue and follow the corresponding comments, and all changes in the manuscript can be tracked by “Track Changes”.
We would like to submit the paper for reconsideration for publication in Sensors. Please feel free to let us know if there are any problems or questions about our paper. Thanks again.
- In general, the revised article has been improved compared with the previous version. Errors such as grammar and symbols in the paper were corrected, and issues raised such as the meaning of each curve and the units of coordinates in Figure 3 were explained and clarified.
Response：
Thank you very much for your patience and positive comments.
- Extensive editing of English language and style required
Response：
Thank you for your comments. We apologize for the inconvenience caused to your review due to our English expression. For your proposed “Extensive editing of English language and style required”, We invited a professional English translator, who guided us through a detailed review, revision and touch-up of the article content. Our revision work includes basic syntax error correction, statement expression replacement, statement logic optimization, etc.
- Basic syntax error correction:
In Line 10 of the Abstract on Page 1, “Moreover, the detection effect of our method was better than that of the YOLOv5 detection model and other orientation detection models.”
- Statement expression replacement:
In Line 3 of the Introduction on Page 1, “In this case, fast and accurate detection of insulators and their defects has become an essential task to ensure the safety of transmission lines [2-3].”
- Statement logic optimization:
In Line 2 of the Abstract on Page 1, “To solve the problem of background interference and overlap caused by the axis-aligned bounding boxes in the tilting insulator detection tasks, we construct an improved detection architecture according to the scale and tilt features of the insulators from several perspectives such as bounding box representation, loss function, and anchor box construction.”
Since there are too many changes, we apologize for not listing the scattered revisions at there one by one. You can track any of our changes in the uploaded word file “sensors-2073401 (Track Changes)”. We hope that these revisions will meet with your approval.
Thank you again for your recognition of our last revision.
Author Response File: Author Response.pdf
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
Point 1: The author should clarify the innovation points of this article. The YOLOv5 rotation frame algorithm and anchor prior are both existing methods. The author should add citations. E.g. 3.2 Scale constraint analysis. The content of this paper is highly similar to the following papers:
(翟永杰,杨旭.结合尺度约束与空间信息的输电线路多金具检测方法[J].华北电力大学学报(自然科学版),2022,49(05):93-100.)
Point 2: The input picture in Figure 1 should be consistent with the output picture. Figure 1 should be placed in Materials and Methods.
Point 3: There is a problem with the position of the detection frame and label in the visualization results of Figure 6. Labels and detection boxs are separated. I doubt the authenticity of the detection results.
Point 4: The comparison of different methods in Table 3 should add visualization results.
Point 5: In 3.1 Test data and parameter settings, The author should add the calculation formula of P and R.
Comments for author File: Comments.pdf
Reviewer 2 Report
To solve the problem of background interference and overlap in insulator identification, this paper proposed a new orientation detection method for tilting insulators, considering angle regression and priori constraints, achieved better results. The research idea is clear and informative. However, I think there are still some problems in the article, please revise and discuss to improve.
1. There are major problems with language expressions: some have basic language errors, some have repetitive or unclear expressions, inconsistent expressions, etc. Please check the full text and correct the relevant errors.
2. Pay attention to the font format of the symbols appearing in the text and in the formulas.
3. Where does the data in Figure 3 come from? Note the meaning of the horizontal coordinates and the units. What are the meanings of the histograms with different color depths and the red curve respectively? Please explain.
4. The anchor box preset parameters extracted by the existing model do not match the insulator scale characteristic. So this paper introduces the K-means method. However, I think the width-to-height ratio corresponding to the cluster center cannot characterize the overall sample distribution. Please give your opinion and understanding.
5. Is it feasible to find the minimum and maximum values of the width-to-height ratio in the training data and set the width-to-height ratio interval based on these two values?
