Combining Fusion-Based Thresholding and Non-Linear Diffusion for Improved Speckle Noise Mitigation in SAR Images

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper presents an interesting analysis and proposal related to the speckle noise mitigation for synthetic aperture radar (SAR) images, with the aim to preserve the quality of the obtained images. The study proposes a fusion-based method that integrates non-linear transform-based thresholds with the advanced noise reduction techniques. The experimental analysis and performance comparison with the existing methods is done fr two simulated SAR images with different noise variance levels.

The paper has some strong aspects and the contribution seem to exist, but it the rationale of the proposed method should better explained, and the performance comparison with the existing state-of-the-art methods more strictly presented and explained. Also, the presentation quality should be improved. Thus, the authors are suggested to:

- Explain the main rationale and provide some insights into the structure of here proposed method. The connection with the existing methods should be described (i.e. in form of proposed changes, improvements, combining and fusion of existing methods  in order to get the proposed solution).

- In the Introduction section the authors present previously proposed methods. Some sort of classification of these methods prior to their overview should be beneficial (a sort of different approaches used in these solutions being defined) which could be connected with the Related concepts section. Also, some sort of discussion in relation to the proposed method could be given (see comment above).

- The performance comparison is done using several filtering methods as the reference methods. However, with the current presentation the reader can not comprehend the connection between  the overview of state-of-the-art methods in Introduction section. The referencing should be applied, or the overview should define the methods using the same names/marks as in the Section 5.

- The authors must provide adequate performance comparison with the all classes of state-of-the-art methods, not just the classical ones, in order to provide correct placement of here proposed method in the current state-of-the-art. Thus, the choice of referent methods used in performance comparison should be explained and justified.

- The technical presentation of tables with the experimental results should be improved - some line numbers are present into the tables, some tables (e.g. Table 7) are badly formatted (overlapped with other tables), etc.

- The technical presentation of related works, equations and formulas, should be improved (e.g. more formal approach to define variables should be used instead x = description under the equations), the same marks for different variables in different parts of the papers should be avoided and correct caps used as in equations, the proper referencing for the prior equations should be given, ...

- A definition of all abbreviations must be given prior to their usage, the references should be given in the sequence of the appearance in the text, etc. Some language mistakes and typos should be corrected.

Comments on the Quality of English Language

The Quality of English Language is acceptable. The final recheck and correction of typos, formatting and sparse language errors should be performed.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The study is interesting and critical considering on the proposes a fusion-based method that integrates nonlinear transform-based thresholding with advanced noise reduction techniques. The proposed method is implemented on two simulated SAR images at noise variance levels of Different performance metrics are considered to investigate the effectiveness of the proposed method. both widely used in the literature. The manuscript is comprehensive and can be accepted after major corrections.

1.      Abstract: One sentence is enough for the motivation of this study. Directly go to the proposed methods with scientific specifications of the remote sensing devices toward improvement nonlinear transform-based thresholding with advanced noise reduction techniques.

2.      Your introduction suffers from a lack of references. Please, expand your reference list. At this point is very poor. You more than 20 references to support your introduction. This is a critical issue to provide to the reader the necessary bibliography towards supporting your methodology.

·         In the line 35 please add reference

·         Line 37 please add reference

·         Line 39-40 please add reference

·         In Lines 53-57-59 please add reference

·         In last paragraph of the introduction please include your objectives.

3.      Methodology

·         Please add section for methodology

·         Please describe your method and originality of your work.

·         Please describe your Algorithm using flowchart  

·         Literature on SAR images

4.      Results

·         Please separate Results and Discussions

·         In the discussion parte please compare your finding with others which was already have done and highlight finding how it is bather then the works which have been done before

·         Please increase the regulation of the graphs in figures 6 also if it is possible use python and plote it 300 dpi

5.      Line 539 Conclusion

 In this section required more improvement, you should present briefly your final results as stated also in your abstract. You should add more information on the limitations you faced during the application of the proposed methodology. Additionally, ways to overcome these obstacles/limitations need to be specified along with the future research you are planning to conduct to improve your results further. The overall conclusion section is poor and needs significant improvement.

·         How can image quality be enhanced while minimizing the loss of key image features and fine details in noise reduction and post-processing steps?

·         Many traditional despeckling methods focus on noise reduction without adequately preserving fine image details. How do you ensure the proposed method will work? usually, this method focuses on edge and texture. How you ensure that this method can adequately preserving fine image detailsthe?

·         The proposed method aims to preserve fine details (like edges and textures) in SAR images while reducing speckle noise. This balance between despeckling and detail preservation is a significant challenge.  How does this research address a specific gap where traditional methods can not fall it can solve?

·         How can speckle noise in SAR images be effectively reduced while preserving fine details such as edges, textures, and object boundaries to improve image quality and post-processing tasks like object detection?

·         for improvement of this work it highly required to Compare the proposed method with a wider range of state-of-the-art despeckling techniques, including both traditional and advanced methods.

·         About calculation is mentioned in the comment which I have uploaded

·         Please correct the reference 33 & 34 same like bellow that I have corrected.

When you use to reference the website please add the accessed data as a bellow.

[1]      “AgIntegrated partners with SkyWatch to bring Earth observation satellite data to the agriculture industry – Agrigate Global.”https://agrigateglobal.com/reads/news/business/agintegrated-partners-with-skywatch-to-bring-earth-observation-satellite-data-to-the-agriculture-industry/ (accessed Sep. 06, 2024).

 

 

Comments on the Quality of English Language

Some minor editing of the English language is required.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The previously presented commands have been modified and the corrections have been made enouph for this manuscript.