Spaceborne Synthetic Aperture Radar Aerial Moving Target Detection Based on Two-Dimensional Velocity Search

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In the reviewed manuscript a method of detecting and focusing on a spaceborne SAR target based on a two-dimensional velocity search is proposed by combining the BP algorithm. The second section presents echo pre-processing and velocity search in detail. Section 3 presents simulation results and analysis. The manuscript is very well prepared. However, I have a few observations:

Comment 1 - Not all abbreviations have been expanded (BP algorithm, PRF, PRT). tr in eq. (4) is not explained

Comment 2 - The figure 3 shows t_a=(M-1)*PRT and the text below it, in line 106 is: "ta=0*PRF to position t_a=(M-1)*PRF". It should be PRT, as t_a is time not frequency.

Comment 3 - in section 3 there is no information what simulation environment was used.

As the above comments do not have a significant impact on the quality of the manuscript, I recommend: Accept after minor revision (corrections to minor methodological errors and text editing)

Author Response

Reviewer 1 Comment 1.

Not all abbreviations have been expanded (BP algorithm, PRF, PRT). tr in eq. (4) is not explained.

Response:

Thank you for the comment. According to your suggestion. "PRT" is the pulse repetition time. “tr” in eq. (4) is range time. We have provided the extension of all abbreviations, such as follows.

Reviewer 1 Comment 2.

The figure 3 shows ta=(M-1)*PRT and the text below it, in line 106 is: "ta=0*PRF to position ta=(M-1)*PRF". It should be PRT, as ta is time not frequency.

Response:

Thank you for the comment. According to your suggestion, we have modified the relevant content, such as follows.

Reviewer 1 Comment 3.

In section 3 there is no information what simulation environment was used.

Response:

   Thank you for the comment. This part uses Matlab 2021a simulation software to simulate.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors apply synthetic aperture radar to detect moving targets on land and water. Their novelty is to detect moving targets in the air.  The difficulty is that targets move in the air faster than on land and sea. The authors suggest a method using spaceborne SARs based on two-dimensional velocity search. The apply clutter suppression and Shannon entropy under different velocity groups to find the velocity most close to the actual velocity of the target.

I recommend this manuscript for publication after language editing and correction of some problems, which are difficult to understand for wide audience. 

 

(1) Please explain what is BP algorithm before using it for the first time.

(2) Please explain what is Shannon entropy. Give a reference. How do you find minimum of this entropy?

(3) How do you apply clutter suppression and clutter suppression methods? Isn’t it the same (line 9).

 

Comments on the Quality of English Language

I recommend publication after minor revision 

Author Response

 Comment 1.

Please explain what is BP algorithm before using it for the first time.

Response:

Thank you for the comment. Back Projection Algorithm (BPA) is a classical algorithm for image processing of time domain echo data, which can accurately construct the target image of the received echo signal in any imaging scene. The imaging algorithm is simple, robust, high resolution, suitable for any orbit or flight path model, there is no oblique approximation assumption, and the motion compensation is easy. Especially suitable for double-base and multi-base SAR imaging. The disadvantage of the BPA is that the time domain imaging method is large and time-consuming, so this paper aims to reduce the amount of computation as much as possible to reduce the running time.

Comment 2.

Please explain what is Shannon entropy. Give a reference. How do you find minimum of this entropy?

Response:

Thank you for the comment. The high-speed movement of the aerial target does not change the total energy of the signal, it will change the transverse energy distribution of the original image, that is, the defocusing phenomenon. Therefore, various image evaluation functions are intuitively used to measure the degree of convergence and dispersion of image energy. In this paper, the minimum Shannon entropy criterion, which is included in the entropy function class of conventional image evaluation function, is selected. Shannon Entropy is an important concept in information theory, proposed by Claude Shannon in 1948, to measure the uncertainty or amount of information in a system. Shannon entropy is often used to describe the degree of uncertainty of a random variable, and can also be used to measure the average information content of an information source. The lower the Shannon entropy, the more stable the system or the image. In this paper, the content to be evaluated is SAR images under different search speed groups, and the Shannon entropy of the current image is calculated and stored according to the order of distance and orientation in different search speed groups. The image is drawn according to the calculated Shannon entropy, and the search speed group with the lowest Shannon entropy is selected. References 31, 32 contain the above.

Comment 3.

 How do you apply clutter suppression and clutter suppression methods? Isn’t it the same (line 9).

Response:

Thank you for the comment. According to your suggestion, we have modified the relevant content, such as the revised article.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This paper proposed an aerial moving target search method based on a spaceborne SAR system in different detection environments. This is an innovative paper, but some information should be added for a better explanation. The following questions should be carefully considered. Specific recommendations are as follows:

1. This paper combines the BP algorithm with the spaceborne SAR system. Have other classical imaging algorithms been tried? And if so, what are the advantages of the BP algorithm?

2. In this paper, Shannon entropy is taken as an evaluation metric. What are the advantages of using Shannon entropy as an evaluation index compared with other methods?

3. Can you provide more information about the connected domain algorithm？

4. Page 4, Fig.3, the coordinate axis information in the figure is not explained in the paper. Please provide a detailed explanation of all the coordinate axes in the figure.

 

 

 

Comments on the Quality of English Language

NAN

Author Response

Comment 1.

This paper combines the BP algorithm with the spaceborne SAR system. Have other classical imaging algorithms been tried? And if so, what are the advantages of the BP algorithm?

Response 1:

Thank you for the comment. The classical imaging algorithms for radar include the RDA in addition to the BPA used in this paper, which is characterized by high resolution, high robustness, simplicity, and high efficiency. However, the RDA also has certain limitations, for the distance and speed close to the target, there may be a blurring phenomenon, and easy to be affected by strong interference sources. At the same time, the RDA has inherent defects, i.e., the distance unit migration correction needs to be interpolated, which not only increases the computational burden of the algorithm, but also reduces the imaging resolution of the algorithm, and the computational amount of the algorithm increases dramatically with the increase of distance migration. Especially for the strabismic SAR working mode, to achieve high precision imaging, its imaging quality is significantly reduced. While the BPA is simple, robust, high resolution, applicable to any orbit or flight trajectory model, has no oblique distance approximation assumption, and easy motion compensation, especially suitable for dual- and multi-base SAR imaging. Therefore, the BP algorithm is chosen for imaging in this paper.

Comment 2.

 In this paper, Shannon entropy is taken as an evaluation metric. What are the advantages of using Shannon entropy as an evaluation index compared with other methods?

Response 2:

Thank you for the comment. In this paper, the minimum Shannon entropy criterion is used to select the most suitable search speed group. Due to the high-speed movement of the aerial target, defocusing will occur in the imaging process, and a single evaluation index is not robust, no matter the peak power, average power, or the combination of the two, it is impossible to evaluate and analyze the image accurately. Compared with the above indexes, the minimum Shannon entropy criterion is more universal and the result is more accurate.

Comment 3.

Can you provide more information about the connected domain algorithm？

Response 3:

Thank you for the comment. Image connected domain refers to the connected region of adjacent pixels with the same pixel value (usually black or white) in an image. An image can contain multiple different connected domains, each of which can be represented as an independent object or region. In image processing, through the connected domain analysis of the image, object detection, target tracking, image segmentation, and other operations can be carried out.

Comment 4.

Page 4, Fig.3, the coordinate axis information in the figure is not explained in the paper. Please provide a detailed explanation of all the coordinate axes in the figure.

Response 4:

Thank you for the comment. The axis perpendicular to the radar track in the horizontal plane is set as the range axis, and the axis parallel to the radar track is set as the azimuth axis.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The specific comments are as follows.

 

1.       In Page 11, it is mentioned that measured SAR image of a region by Gaofen-3 is selected as the background for echo simulation. I wonder how to obtain the radar echo of the composite scene consist of the background and target?

2.       It will be good to separate “Discussion” part.

3.       How to evaluate the performance of the proposed scheme in this paper? It will be good to add quantitative evaluation indicators?

4.       In Fig. 4, how to deal with Doppler effect arising sea wave movement?

5.       Page 4，the first sentence in the last paragraph, “In the figure above”, the Figure No. should be specified clearly.

6.       “2.1.1. Ground Detection Echo Pre-processing” may be modified as “2.1.1. Echo Pre-processing for Ground Detection”

7.       “2.1.2.Sea Detection Echo Pre-processing” may be modified as “2.1.2. Echo Pre-processing for Sea Detection”

8.       English writing and grammar should be checked and improved.

9.       New References relevant to this research are insufficient. Several References cited in this manuscript are Dissertation in Chinese. It seems not appropriate and also hard for non-Chinese readers.

Comments on the Quality of English Language

English writing and grammar should be checked and improved.

Author Response

Comments 1.

In Page 11, it is mentioned that measured SAR image of a region by Gaofen-3 is selected as the background for echo simulation. I wonder how to obtain the radar echo of the composite scene consist of the background and target?

Response 1:

Thank you for the comment.The measured data obtained by Gaofen-3 are used as background clutter, and on this basis, aerial moving targets and noise are added by simulation.

Comments 2.

It will be good to separate “Discussion” part.

Response 2:

Thank you for the comment. According to your suggestion, we have modified the relevant content, such as the revised article.

Comments 3.

How to evaluate the performance of the proposed scheme in this paper? It will be good to add quantitative evaluation indicators?

Response 3:

Thank you for the comment. The strengths and weaknesses of the scheme can be seen by comparing the approximate profiles of the preset target imaging with the carryover search velocity imaging. The average error of the magnitude on the corresponding pixel points is used as a quantitative evaluation metric. Analyzed from the perspective of each pixel point amplitude, the average error is 4.1512%.

Comments 4.

In Fig. 4, how to deal with Doppler effect arising sea wave movement?

Response 4:

Thank you for the comment. Unfortunately, due to the high-resolution characteristics of SAR, the Doppler effect arising sea wave movement has little effect on aerial moving targets, so it is not considered in this paper. This part of the work will be carried out in the follow-up research.

Comments 5.

Page 4，the first sentence in the last paragraph, “In the figure above”, the Figure No. should be specified clearly.

Response 5:

Thank you for the comment. According to your suggestion, we have modified the relevant content, such as the revised article.

Comments 6.

“2.1.1. Ground Detection Echo Pre-processing” may be modified as “2.1.1. Echo Pre-processing for Ground Detection”

Response 6:

Thank you for the comment. According to your suggestion, we have modified the relevant content, such as the revised article.

Comments 7.

 “2.1.2.Sea Detection Echo Pre-processing” may be modified as “2.1.2. Echo Pre-processing for Sea Detection”

Response 7:

Thank you for the comment. According to your suggestion, we have modified the relevant content, such as the revised article.

Comments 8.

 English writing and grammar should be checked and improved.

Response 8:

Thank you for the comment. According to your suggestion, we have modified the relevant content, such as the revised article.

Comments 9.

 New References relevant to this research are insufficient. Several References cited in this manuscript are Dissertation in Chinese. It seems not appropriate and also hard for non-Chinese readers.

Response 9:

Thank you for the comment. According to your suggestion, we have modified the relevant content, such as the revised article.

Author Response File: Author Response.pdf