Analysis of the Matchability of Reference Imagery for Aircraft Based on Regional Scene Perception
Round 1
Reviewer 1 Report (Previous Reviewer 1)
This paper proposes an innovative method for matchability analysis of reference imagery based on regional scene perception. I have re-reviewed the revised version of this paper with reference to the comments of other reviewers in the previous round.
The authors have revised the questions and answered my queries I raised in the previous round. I found that the authors have reorganized the structure of the paper and added more evaluation metrics to illustrate the strengths of this article's methodology.I have no further suggestions for revision.
The paper is now significantly improved in terms of completeness of writing, innovation of methodology, and adequacy of experiments. Therefore, I believe that the current version fulfils the requirements for publication in Remote Sensing.
Author Response
Thank you very much for your work and advice on the paper, which made this paperperform even better.
Reviewer 2 Report (Previous Reviewer 2)
no further comments
Author Response
Thank you very much for your work and advice on the paper, which made this paper perform even better.
Reviewer 3 Report (Previous Reviewer 3)
- In Page 2 Line 50, change … (b) are indicating … -----> (b) indicates …
-Redraw the figure 5 (which shows the structure of the RSU Module) to appear in a better resolution.
- Redraw the figures 10 and 12 (Figures 11 and 13 in the previous version ) to appear in a bigger font size of both the x-axis and y-axis. Notice: you already modified in the file named " remotesensing-2596937 author's reply".
- In page 2 line 78 , the word " conventional" was repeated two times, remove one.
- In page 2 line 79 , the word "Networks" was repeated two times, remove one.
-(b) indicating areas with similar features ...> (b) indicates areas ----
-Please, recheck again the whole manuscript for any language mistakes
Author Response
First of all, thank you very much for your review work and suggestions to favorably enhance the presentation of the paper.
1、According to your suggestion, I have revised the incorrectly used words and repetitive words in the paper that you pointed out. And I have rechecked and revised the language mistakes in the article.
2、We have redrawn Figure 5 to make it more presentable and higher resolution.
3、We have redrawn Figures 10 and 12 to make them more clearly visible.
Reviewer 4 Report (New Reviewer)
The problem of the correct positioning of the object presented by the authors is justified from the point of view of the use of flying objects (aircraft). The idea is good, but it will be credible based on the key elements of the unchanging environment (fixed infrastructure points of large cities). my doubt concerns: how the authors will secure the process of positioning objects in areas of changing agricultural crops.
The literature is correctly selected for the issue under consideration. The research workshop is correct and the mathematical side corresponds to modern scientific studies. I am asking the authors for a laconic explanation of my doubts. I say well done.
Author Response
First of all, thank you very much for recognizing and validating the work of the article and I am very glad to answer your concerns. This paper innovatively proposes a core approach to quantifying matchability by utilizing scene information from imagery. The core of the work in this paper is to evaluate the matchable area and matchable performance of this predefined reference base map during the working phase of the vehicle, without the situation that the crops keep changing. On the other hand, the network architecture in this paper adopts a modular design that takes into account scene perception. Although the crops keep changing, the changes in their whole scene characteristics are minor, thus enabling the assessment of the matchability performance of the whole scene.
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
This paper presents an innovative approach to quantifying matchability in scene matching for aircraft positioning, and the proposed RSPNet shows promise in improving the success rate of scene matching. The paper combines the transformer and attention mechanism properties that are currently exploding in CV to bring newer hints to the traditional application of matchability analysis. However, there are a few minor issues with the article that need to be revised, and it is recommended that it be revised for publication.
1. The reference images in the text are named repetitively, such as "reference benchmark images", "reference images", "benchmark images". Please refer to related literature to modify or unify them into a fixed proper name.
2. Please explain the legend of the blue color grading from 1 to 10 in Figure 1.
3、In section 3, there are three red boxes given in Figure 2, which do not correspond to the range of areas given in the text, please confirm.
4, There is an extra horizontal line below the graph in Fig. 7, please make sure that the problem exists in other graphs and modify it.
5、The method of this paper proposes to take into account the scene perception, please explain how it is reflected in the method of this paper.
6、What is the range of output of a grading in the result graph of image matching ability predicted by the method of this paper, and why? Can it be output pixel by pixel?
7. Please confirm whether there is any problem with the format of the references.
Please pay attention to the standardization of the writing.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
After carefully reading this article, I have not found any valuable innovations. The three main innovative contributions mentioned by the author are not valid. What are the differences, advantages, and uniqueness of the dataset? Is there any value in making it public? What are the differences in quality and performance compared to existing datasets? If the data itself does not involve original innovation, the author cannot classify it as a major contribution. The RSPnet mentioned by the author is not innovative, as most of the DL related papers that have been published currently have saliency analysis modules and uniqueness analysis modules In addition, the author has only made minor adjustments to the existing net and lacks innovation. Figure 8 is no different from existing multi-level methods, and the author cannot plagiarize the results of others. The evaluation criteria provided by the author are too simple, and these basic indicators are difficult to reflect unique advantages. It is recommended to add indicators such as ROC. The comparison methods listed in Table 1 have no practical value and can only be referred to as the performance of methods under different functional effects, rather than a comparison of advanced algorithms.
major problems on the language
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
In general, the form and the analyses of the problem are correct. The paper has a form of original article. This work deals with the Analysis of the matchability of reference imagery for aircraft based on regional scene perception .
This manuscript can be accepted for publications after some minor corrections and modifications as given:-
- In page 2 line 49 , change (a) is indicates…. -----> (a) indicates….
- Similarly, in line 50, change … (b) are indicating … -----> (b) indicates …
- For more clarification, it is better to redraw figure. 2 to appear in a better resolution with a bigger font size of both the x-axis and y-axis. It is also important to write the symbols a, b and c on its place of the red box in the graph.
- In page 6 line 49, you wrote “The perturbations contain color, greyscale distortion, Gaussian, pretzel, speckle noise and ………
What do you mean by the speckle noise? It is better to give a reference.
- For more clarification, it is better to redraw the figures 4 and 6 to appear in a better resolution.
- Redraw figures 7, 8, 9, 10, 12 and 14 to appear more clearly or in a bigger font size.
- Redraw the figures 11 and 13 to appear in a bigger font size of both the x-axis and y-axis. Also to appear in a frame with major and minor grids.
- In page 10, line 334, Change … B_4 ….. ----> B4 …..
- In page 14 line 445, change …. Where A, B, C and D are … -----> Where (a), (b), (c) and (d) are….
- In Page 18 line 550 Please add the words Conventional Neural Network (CNN)
- In table 2 and table 3, change R2 to R2
- It is better to write the results you got from your method to support the conclusion section.
There are some minor comments about the language, such as the comments which i mentioned in page 2 line 49 and line 50.
So it is better to check the whole manuscript for the language.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
It is obvious that the author did not make any modifications or provide a satisfactory response based on the review comments. 1. There was no more profound explanation of innovation, and the differences between the methods used by the author and existing methods were not significant. There was no improvement in the theoretical content. 2. The author did not conduct more extensive indicator validation and evaluation, which could not confirm the advantages of the method. In summary, this article cannot meet the publication requirements.
should be improved
