The Application of CNN-Based Image Segmentation for Tracking Coastal Erosion and Post-Storm Recovery

Round 1

Reviewer 1 Report

Comments for author File: Comments.pdf

Author Response

I have attached the file with the response for your convenience. Thank you.

Author Response File: Author Response.docx

Reviewer 2 Report

Please see the attachment.

Comments for author File: Comments.pdf

Author Response

I have attached the file with the response for your convenience. Thank you.

Author Response File: Author Response.docx

Reviewer 3 Report

The work is well done is very clear, I have no notes for the authors

Author Response

Reviewer 4 Report

This paper presents a camera-based approach for detection of post-storm beach recovery by applying CNN classification. The topic is interesting, and the use of on-site camera is a very practical while novel choice.

However there is one issue I would hope the authors further address to. In processing camera images, weather conditions could be a major complication. In the paper, the "blurry" or "haze" images, also the "high luminance problems", are probably the consequence of specific weather conditions. As I understand, it is not easy to realize a classifier as such to adapt to varying shoreside weather. 

Do you think it is necessary to find a systemic approach to deal with this problem? Maybe train models for each weather condition and introduce weather station data & camera position status to determine which model to use?

Also, in the pdf file I acquired, the resolution of Figures 4~9 are not high enough. These are very important images in understanding the model validation results, so please ensure sufficient quality.

Author Response

I have attached the file with the response for your convenience. Thank you.

Author Response File: Author Response.docx

Reviewer 5 Report

Review Report for the Manuscript “The application of CNN-based image segmentation for tracking coastal erosion and post-storm recovery”

Introduction is too short to introduce the importance of Machine learning for coastal monitoring and assessment, an extensive analysis of literature review is required.

In the methods section:

It is not clear how authors used CNN techniques, algorithms and software.

Specification of Dataset and uncertainty assessment

Most of figures without scales, definitions and they are not clearly display.

Results still not clear.

The novel methodology applied in this study needs to more clarification and analyses.

Results refer to only image classification using segmentation techniques for spatiotemporal images pre and after the event, the tracking coastal erosion and inundation still not clear and to more work.

Comments for author File: Comments.pdf

Author Response

I have attached the file with the response for your convenience. Thank you.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

This paper has greatly improved. The authors should add more state-of-the-art work as compared methods, especially the work published in the last three years. The Tables 3-5 have been removed, leading to the lack of quantitative comparison.

Author Response

I have attached the file with the response for your convenience. Thank you.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors considered all my suggestions. I strongly believe that the manuscript is now much more comprehensive and clear. Many changes have been added to the manuscript. I now suggest that the authors pay attention to the section numbers, font type and size, citation, etc. to maintain clarity and standardization.

Minor review

"Results and Discussion" instead of "Results and Discussions".

Keep the first letter in uppercase for each word of the section and subsection titles.

Author Response

I have attached the file with the response for your convenience. Thank you.

Author Response File: Author Response.pdf

Reviewer 5 Report

Second Review Report for manuscript: The application of CNN-based image segmentation for tracking coastal erosion and post-storm recovery

(IoU) abbreviation should add in the abstract

Page 4: 133-140 need more details

Author should clarify how he distinguish between sand deposits

Figure 2. Stages and source are required.

 Reference 18. Citation missed inside the text.

In figure 4: the color of the legend does not the same inside the Figure; what is the “Object”

The same Equations in the Page 10 have inserted in Page 7.

References should be the same font, size and system (see Reference 2, and 3)

Comments for author File: Comments.docx

Author Response

I have attached the file with the response for your convenience. Thank you.

Author Response File: Author Response.pdf