Infrared Small Marine Target Detection Based on Spatiotemporal Dynamics Analysis

Round 1

Reviewer 1 Report

Dear authors!

The paper is dedicated to an interesting and valuable problem of Small Marine Target Detection Based on Spatiotemporal Chaos Analysis. The paper is well-organized and the obtained results clearly show that the advantage of the proposed approach.

Nevertheless, several remarks should be made.

1. The English language needs improvement.

2. Please, clarify the used of word "chaos". By definition, chaos is a nonlinear dynamical process with positive largest Lyapunov exponent. In your research, as I can judge, existence of chaos is not necessary, so would the word "dynamics" instead of "chaos" more strict?

3.  There is lack of substantiation, why authors prefer using RBF ANN over the other types of dynamic reconstruction algorithm. I suggest including a brief review of alternative approaches with justification why RBF ANN is selected. In this review the following work are proposed to be considered:

"Algebraic method for the reconstruction of partially observed nonlinear systems using differential and integral embedding",

"Noise-tolerant algebraic method for reconstruction of nonlinear dynamical systems",

"Identifying empirical equations of chaotic circuit from data".

4. At line 354 you say, that in spatial domain the saturation is reached for m = 7, but I would argue that the value m = 12 is closer to truth. How the saturation is formally determined?

5. It is not very clear which data is used for training the ANN: the entire frame or a certain part of it? What are computational costs of the algorithm?

6. It would be interesting to see an image with a target and an image reconstructed by the ANN, not only residual images. Are these entire images or fragments of a certain larger image?

7. I suppose, that comparison with only top-hat filter might be insufficient since this filter is very simple and a number of alternative approach are known. Please give a comparison with any one more other alternative approach or give references where top-hat filter is proven to be the best filter among the others. I recommend comparing with α-β-γ filter or spatio-temporal domain joint filter. See references

https://doi.org/10.3390/e24020250

http://dx.doi.org/10.1109/ICOSP.2008.4697627

Author Response

Dear reviewer,

Thank you very much for your letter and for the reviewer’s comments concerning our manuscript entitled Infrared Small Marine Target Detection Based on Spatiotemporal Dynamics Analysis. The original title is Infrared Small Marine Target Detection Based on Spatiotemporal Chaos Analysis. According to the suggestion, we replace the “chaos’’ with the “dynamics”. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied the comments carefully and have made correction which we hope meet with approval. The important revised places are highlighted with yellow background in the manuscript. The details are written in the attached letter.

Kind regards,

Li Zhengzhou

Author Response File: Author Response.docx

Reviewer 2 Report

This paper has proposed a spatiotemporal chaos analysis-based marine target detection algorithm to enhance the performances of infrared small target detection with sea clutter suppression. The topic is interesting, and the manuscript is well organized. Whereas, I still have the following concerns.

1. Contributions of the manuscript should be further highlighted in introduction and conclusion.

2. Section 2.1 and 2.2 is named “data selection”. The authors do not select data from a data set. The process is more like data or feature extraction. Please reconsider the name.

3. In section 3.1, the model is defined by equations. It is better to explain the thought in the begging of this subsection, making the model easier to understand.

4. In section 5.1, The authors claim “Figure 6(d) is used to verify it’s chaotic in spatiotemporal domain”. Why do not use other subfigures in Figure 6?

5. In Figure 7, “the spatial delay time and the temporal delay time are 4 pixels and 5 frames”. How to judge the spatial delay time and the temporal delay time from the figure?

Author Response

Dear reviewer,

Thank you very much for your letter and for the reviewer’s comments concerning our manuscript entitled Infrared Small Marine Target Detection Based on Spatiotemporal Dynamics Analysis. The original title is Infrared Small Marine Target Detection Based on Spatiotemporal Chaos Analysis. According to the suggestion, we replace the “chaos’’ with the “dynamics”. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied the comments carefully and have made correction which we hope meet with approval. The important revised places are highlighted with yellow background in the manuscript. The details are written in the attached letter.

Kind regards,

Li Zhengzhou

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript presents a nove spatiotemporal chaos analysis-based marine target detection algorithm to enhance the performances of sea clutter suppression and infrared small target detection. The following questions need to be addressed.

(1) What is the robustness or generalization of the proposed algorithm, such as in low SNR case and at the conditions of different scenes？

（2) The computational complexity and the running time of the proposed algorithm need to be provided, together with comparation with other algorithms.

(3) What is the major limitation of proposed algorithm, which can be addressed in the conclusion.

(4) The English usage, Equation and figures need to be carefully checked.

Author Response

Dear reviewer,

Thank you very much for your letter and for the reviewer’s comments concerning our manuscript entitled Infrared Small Marine Target Detection Based on Spatiotemporal Dynamics Analysis. The original title is Infrared Small Marine Target Detection Based on Spatiotemporal Chaos Analysis. According to the suggestion, we replace the “chaos’’ with the “dynamics”. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied the comments carefully and have made correction which we hope meet with approval. The important revised places are highlighted with yellow background in the manuscript. The details are written in the attached letter.

Kind regards,

Li Zhengzhou

Author Response File: Author Response.docx