Gudalur Spectral Target Detection (GST-D): A New Benchmark Dataset and Engineered Material Target Detection in Multi-Platform Remote Sensing Data
Round 1
Reviewer 1 Report
Minor correction on figure 4: spaceborne hyperspectral is spaceborne multispectral
What bands of sentinel 2 were used in the analyses? From figure 19, it appears you used all the bands.
I am concerned about two serious omissions in your analyses and discussions:
- the adjacency effect. This results in signal from adjacent material to be included in the measured reflectance of the target for a certain distance inward from the boundary. This contamination can be significant, especially if the atmosphere is not perfectly transparent. You must account for this in your analyses for the airborne and spaceborne sensors.
- The problem with spatial resolution of your sensors in relation to the 10x10 m targets. For Sentinel-2 data, you will NEVER get a perfect alignment of a pixel over your target. EVERY pixel will be mixed, contaminated with the adjacent background. This must be thoroughly discussed. It probably means that all of your Sentinel analyses are wrong or meaningless.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The research paper is very interesting and very well written.
About the resported accuracy 80%, is relatively low since there is other studies in the field reporting higher accuracy. Did the authors tried some signal processing techniques before applying the proposed target detection algorithm, to further enhance the results.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
In this paper, the authors acquired a benchmark multi-platform hyperspectral and multispectral remote sensing dataset named ‘Gudalur Spectral Target 21 Detection (GST-D)’ dataset. Then, statistical and subspace detection algorithms have been applied to the benchmark dataset for the detection of targets. They finally validated the target detection results using the receiver operation curve (ROC). Generally, the proposed idea is very interesting, however, some revisions have to be made and some parts of the manuscript not complete to claim the advantage of the proposed models.
1) Could the authors explain what is the main contribution of the proposed idea than over existing approaches?
2) Could the authors explain what is the motivation to use the receiver operation curve (ROC)?
3) Could the authors add more clarifications about the detection score imagery in figure 4?
4) The English and format of this manuscript should be checked very carefully.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
thank you for your responses
