Next Article in Journal
Cooling by Cyprus Lows of Surface and Epilimnion Water in Subtropical Lake Kinneret in Rainy Seasons
Next Article in Special Issue
An Innovative Supervised Classification Algorithm for PolSAR Image Based on Mixture Model and MRF
Previous Article in Journal
Heterogeneity of Increases in Net Primary Production under Intensified Human Activity and Climate Variability on the Loess Plateau of China
Previous Article in Special Issue
Built-Up Area Extraction from GF-3 SAR Data Based on a Dual-Attention Transformer Model
 
 
Article
Peer-Review Record

Water Surface Mapping from Sentinel-1 Imagery Based on Attention-UNet3+: A Case Study of Poyang Lake Region

Remote Sens. 2022, 14(19), 4708; https://doi.org/10.3390/rs14194708
by Chaowei Jiang 1,2,3, Hong Zhang 1,2,3, Chao Wang 1,2,3, Ji Ge 1,2,3 and Fan Wu 1,2,*
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Remote Sens. 2022, 14(19), 4708; https://doi.org/10.3390/rs14194708
Submission received: 22 August 2022 / Revised: 15 September 2022 / Accepted: 16 September 2022 / Published: 21 September 2022
(This article belongs to the Special Issue SAR in Big Data Era II)

Round 1

Reviewer 1 Report

The paper is generally interesting and meaningful in this field. It develops an Attention-UNet3+ model with SAR images for water surface mapping, which extracts and utilizes the full-scale features of the input images through the encoding-decoding structure. The experiment results show that the proposed method outperforms the conventional threshold segmentation and other deep learning models. The manuscript is overall well organized, and the results are appropriately presented. However, I have a number of concerns before this study can be further considered for publication. Several comments are given as follows.

1. In the introduction part, the existing research is not well investigated and summarized, e.g., the current research status of full-scale skip connections and attention mechanisms.

2. In fact, the ‘Attention U-Net architecture’ has been proposed for medical image segmentation. However, authors use the attention module with the same structure as the existing literature. It seems that the proposed method is lack of innovation. It is suggested to highlight the innovation of the proposed method.

3. In the main contributions of the research, it is mentioned that ” The staged output of the decoder is used to improve the model efficiency, which enables the model to have fast segmentation capabilities”. The demonstration is insufficient. Authors should provide some data metrics for the model efficiency of the proposed method and compare it with other models to verify the superiority of the method.

 

4. The language could be polished by an English native speaker.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

Dear authors,  

The effort of water surface mapping using SAR data and deep learning is interesting and would bring a significant contribution in this field.  

Manuscript is being well written, but more addition information about using mNDWI should be provided to clarify it use instead of field sampling or another water indices.  

I wish that my comment would be helpful in improving the quality of this research.  

Thank you.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Back to TopTop