Detection of Black and Odorous Water in Gaofen-2 Remote Sensing Images Using the Modified DeepLabv3+ Model

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The article demonstrates a commendable level of writing, and the findings exhibit promise. However, there are areas that could be enhanced.

In the introduction (specifically between lines 108-116), I recommend refraining from introducing specific results. Instead, focus on emphasizing the study's significance, its rationale, research inquiries, and novel contributions distinct from prior work.

I propose a dedicated section solely for presenting the results, separate from the "4. Experiment and Evaluation" section. This distinct segment would allow for a clear demarcation of the experimental study and the results themselves.

A critical addition would be a comprehensive discussion of the research outcomes, including a comparative analysis among algorithms. This comparative aspect is vital to offering a detailed interpretation of the findings concerning previous achievements in the same research domain.

Concluding the study with a section dedicated to its limitations could provide valuable insights into the future orientation within this research field.

Author Response

Please refer to the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Sir/Madam,

The paper entitled “Detection of Black and Odorous Water in Gaofen-2 Remote Sensing Images Using the Modified DeepLabv3+ Model” submitted by Huang J. et al. presents an excellent method to identify BOW that overwhelms existing methods in many aspects. The authors also explained the comprehensive study background including related works investigation and positioning methods of BOW in a detailed way. In addition, they also demonstrated the performance of the proposed method by analyzing real data.

In this regard, the paper provides us with very useful information. It is therefore suitable for publication with its present format in every aspect of article formulation except some revision recommendations which are written on a separate page.

 Best regards,

 

Reviewer

 

 

 

Additional comments:
Abstract
  Line 2-4: High spatial resolution of RS images In general, high-resolution images can extract target features better than low-resolution images. Therefore, I think it is appropriate to change the expression of this sentence as follows: … the high spatial resolution images often have problems with accurate feature extraction despite their high resolution.   Line 24: RSDB What stands for RSDB? Line 122: 2. Related Work Please add implications briefly related to the background or purpose of this study in the end of this chapter, considering the characteristics, pros and cons of related works, etc.   Line 122: 2. Related Work Please add implications briefly related to the background or purpose of this study in the end of this chapter, considering the characteristics, pros and cons of related works, etc.   Line 262: 4. Experiment and Evaluation Please add a brief introduction to Geofan-2 image somewhere at this beginning part.   Line 326: The reason for accuracy degradation Since all results in Table 1 were conducted under the same conditions, a more persuasive explanation except ignoring the problem of imbalance in the dataset is needed as to why the accuracy decreases especially in OA of the Modified Modified DeepLabv3+ against DeepLabv3+.      Line 346: Figure 6 1)      (c)SeNet should be corrected as (c)SegNet. This also applies to Figure 7. 2)      In the second row of the picture, BOW cannot be recognized in the original image.   Line 536: Accessing the website Please check whether this website can be accessed. I currently can't access this website.. - The End-

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Introduction: please present better the general context of the current issue, what are the limits of previous studies and what is the novelty of your research. Add more bibliographic references.

Phrases must be polished because sometimes the content is not very clear

r. 45 water environment = water body

Add more references: r.55  many research scholars have conducted BOW monitoring ... but no citation.

What are the methods by which an "water body" is odorous except the color?

At the end of section Result present the limit of your research and future directions

r.502-506 move to conclusion

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

This is an interesting paper describing the detection of Water on Gaofen-2 Satellite Images using the DeepLabv3+ Model.

I would like to see this study published; however, in its current form, the paper is not ready for publication.

While the results are fairly well-elaborated, the text requires significant shortening and focusing in multiple areas. Some main points are detailed below.

 

My suggestion is to merge the Sections Methodology and Related Work. This part of the article is very elaborate and represents a significant part of the article. I leave the final decision to the authors. 

 

Section 4.1 Dataset Description: this part needs to be rewritten.

 

Detailed comments:

 

No information about the satellite system used:

- Why were images from this system used?

- Did someone select the images for the test? 

- Were dates and spectral ranges (B4/NIR) taken into account?

- Did the spatial resolution of the chosen system matter?

 

Test area:

- no description of the test area,

- no map showing the location of the test area,

- Was only one city selected (Yantai, China) ?

- Why was the proposed algorithm not tested for other areas? What were the reasons for this?

- Why were several test areas not selected to obtain reliable results?

 

I would like to see more tests to verify the actual effectiveness of the proposed methodology.

A Discussion Section is missing, where the authors could have compared the results with other studies. This section should be expanded and should be the most interesting part of the submitted article.

The authors focused on the experiment itself, which was not properly planned.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 4 Report

Comments and Suggestions for Authors

The comments have mostly been taken into account in the revised version.