A Destriping Algorithm for SDGSAT-1 Nighttime Light Images Based on Anomaly Detection and Spectral Similarity Restoration
Round 1
Reviewer 1 Report
This paper proposed an algorithm to remove strips from GIU nighttime images. The results, compared with some traditional methods, look promising. The algorithm is clearly presented.
Is the user-defined-threshold (L210) applicable universally or we need to adjust it when processing different image, as far as I understand, the intensity of each pixel will be different for different places, e.g., different cities. What is the typical values.
If the pixels of all the bands is abnormal, e.g., in the strip, how to use the spectral similarity method to restore the pixel value?
Please pay attention to the English, some minor comments are as follows:
L35: please define NTL when it is first used in the article.
L38-40: NTL remote sensing is also used to study the impact…environment because of its high accuracy for depicting artificial light.
L40: please define GIU here.
L40-L42: This sentence should be refined; it is not clear.
L44: have
L66: what is “the other”.
L93: Therefore, previous destriping algorithm may not be …
L28: GIU images are vastly different.
L179: I think you should use “except for” here
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 2 Report
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
In this paper, the author novel destriping algorithm based on anomaly detection and spectral similarity restoration (ADSSR) for the GIU image. Experiments demonstrate the robustness of the ADSSR algorithm, and the experimental results also show that ADSSR algorithm has outstanding performance. And the paper provides a direct and meaningful method to improve the quality of the SDGSAT-1 nighttime light image.
However, there are still small problems in the paper. The comments are as follows:
1. In equation (5), the calculation formula of the Euclidean distance is incorrect.
2. In section 4, the test data and the quality assessment indicators are detailed, but the environment of the experiment and the specific parameters of the ADSSR algorithm are not introduced, which need to be supplemented.
3. Some grammatical errors and clerical errors in the paper need to be corrected.
1) In Figure 4, “Restoration based spectral similarity” should be changed to “Restoration based on spectral similarity”.
2) In line 175, “The details of the three parts are presented in Section 2.1, 2.2, and 2.3, respectively”, is “Section 2.1, 2.2, and 2.3” a typo? Is it “Section 3.1, 3.2, and 3.3”?
3) In line 261, “needs to be unrestored” is confusing, is it “needs to be restored”?
4) In line 420, “The outstanding performance and robustness of the proposed algorithm provide a feasible method to improve the GIU image quality.”, the subject of this sentence is not used properly, is it more reasonable to change the subject to “this article” or “this paper”?
Author Response
Please see the attachment.
Author Response File: Author Response.docx