Feature Selection for Edge Detection in PolSAR Images

Round 1

Reviewer 1 Report

Dear Authors,

The subject addressed in the article is interesting, but you should improve the work as pointed out below:

a) The Abstract is very synthesized, and the authors should improve the description of what was done. For example, better explain why Polarimetric data were used and inform which SAR sensors were used and their bands (L and P), as well as the results achieved. The Abstract needs to be revised.

b) In item “2.1. Data” were presented the data used, however the description of the data should be improved. For example, for SAR data acquired in the L and P bands, the data acquisition dates and by whom they were acquired were missing. In the case of the P band, as it is a commercial sensor operated by a company in Brazil, was the use of this data authorized by the owner of the sensor system or by the contractor for the aerial coverage? No acknowledgment was done in the document related to this.

c) In item “2.2. Ground References and Images”, the authors should explain how these references were obtained. It was not clear in the text and only orange vectors are presented on the images in colored composition (Pauli).

d) In item “3. Results”, the description of the computer used should be present in item “2. Materials and Methods”.

e) In “Figure 7.” It's not clear to me what the authors are trying to show with the overlapping of the orange dots over the different images. It is necessary to improve the explanation in the text. The same comment applies to the other figures that following.

f) Item “5. Discussion and Conclusions” should be divided into “5. Discussion” and “6. Conclusions”. Furthermore, the authors should explain the impact of using different bands (L and P) on obtaining the results.

Finally, I suggest a general review of the document in such a way that the presentation of the data used, the methodology adopted, its results, the discussions based on the results obtained and an objective and clear conclusion is improved.

Author Response

Thank you very much for handling this manuscript.

We have prepared a revised version taking into account all the comments and suggestions made.

In fact, we found the review well-informed and constructive. We would like to thank the reviewer.

This response letter addresses all the comments in red, followed by
our reactions, and, whenever necessary, the changes made.

As a final comment, we would like to stress the link to a repository with the codes and data promote the reproducibility of this work.

The authors

Author Response File: Author Response.pdf

Reviewer 2 Report

In this paper, the Authors propose a new fused methodology to aiming detecting edges within PolSAR data. The method works on the ratio images and several optimization methods are combined.

The paper is certainly well written, sound and convincing.

This reviewer has only two minor issues:

- include computational cost in tables for all methods,

- include a sensitivity analysis. There are several empirically set parameters that are relevant to the performances of the method. 

Author Response

Thank you very much for dealing with this manuscript.

We have prepared a revised version considering all the comments and suggestions made.

In fact, we consider the review to be well-informed and constructive. Significantly, the timetable that showed interesting results scored at the conclusion of the paper. 

This response letter addresses all comments in red, followed by our reactions and, where necessary, our changes.

As a final comment, we would like to stress the link to a repository with the codes and data promote the reproducibility of this work.

The Authors

Author Response File: Author Response.pdf

Reviewer 3 Report

Feature Selection for Edge Detection in PolSAR Images:

·        Lines 80-83 are not necessary and should be deleted.

·        In the last paragraph of the Introduction, the authors should mention the weak point of former works (identification of the gaps) and describe the novelties of the current investigation to justify that the paper deserves to be published in this journal.

·        Discuss more the edge detection using S-ROC fusion.

·        “The methods proposed to obtain the edge evidences using the information fusion (S-ROC, and τS-ROC) work well when applied to images obtained by different sensors: FLEV and SF were obtained with the AIRSAR sensor, S01 and S02 were obtained with the OrbiSAR-2 sensor, and, finally, the LAA and LAM were obtained with the UAVSAR sensor.”. Explain.

·        Focus on the advantages/disadvantages of the proposed method concerning the obtained results.

·        What are the strategies/recommendations to reduce uncertainties in this study?

·        It is necessary to explain the sources of error in this study to consider them in the next investigations.

 

·        At the end of the manuscript, explain the implications and future works considering the outputs of the current study.

Acceptable.

Author Response

Thank you very much for handling this manuscript.

We have prepared a revised version taking into account all the comments and suggestions made.

In fact, we found the review well-informed and constructive. We would like to thank the reviewer.

This response letter addresses all the comments in red, followed by
our reactions, and, whenever necessary, the changes made.

As a final comment, we would like to stress the link to a repository with the codes and data promote the reproducibility of this work.

The authors

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

I appreciate the authors addressing the comments. The manuscript can be accepted in its current form. Congrats!

Acceptable.