Small Recreational Boat Detection Using Sentinel-1 Data for the Monitoring of Recreational Ecosystem Services
Round 1
Reviewer 1 Report
In this paper, a two-stage method is proposed for small ship detection in SAR image. It is very interesting to detect small recreational boats in lakes. However, there are still several problems to be solved:
1: It is recommended to add more information in Table 1 to explain the imaging parameters, such as incident angle.
2: I noted that, some similar references as DOI: 10.1117/1.JRS.14.044522 and DOI: 10.1109/TGRS.2022.3216532 also proposed two-stage methods. Please compare the proposed two-stage method with them in the paper.
3: Since this paper focuses on recreational boat detection, it is better to show some real images of this kind of ships in the paper.
4: Line 55: helps->help. Some English grammar errors need to be further revised.
5: In Section 1, the contributions of this paper should be given more detailedly.
6: Line 185, the authors state that HV was too noisy. Please give the visual images of HV and HH at one same site. If so, readers can better understand this point.
7: Please add some comparison methods to further verify the superiority of the new method.
8: As I know, Touzi give the detailed angle range to obtain the conclusion that co-polarization is the best choice. Please add this angle range in Conclusion.
Author Response
We would like to thank for all comments. Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In general an interesting paper.
Please correct the following:
Line 210: the sentence is unclear. What is the "...basis of the carried out...".?Are the values for the coefficient derived from the investigation. If yes, how was that done?
Lines 228-230: Paragraph needs improvement. Which are the "two kinds of validation"? they are explained in the next paragraphs but it would be nice to identify them from the beginning of the section.
Figure 5: the values and legend are not readable. Maybe separate the left image from the two graphs and zoom in.
Author Response
We would like to thank for all comments. Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper is devoted to detection of the small boat (<10 m in length) with the use of Sentinel-1 SAR images in lake conditions for the monitoring of recreational ecosystem services. The authors developed two-step algorithm for extracting the boats. The method was tested with the use of 14 images of different lakes, namely, Great Masurian Lakes, situated in Poland. The method is demonstrated to be of rather high (maximal - 88.17%) accuracy with correlation of satellite-based map and field observation of 0.76. Expert wants to say that accuracies listed in table 4 are practically the same (76 – 88%). The only remarks are – to avoid abbreviations in abstract and keywords (SAR) and to make heading of table 4 more understandable (what are “PA”, “UA”, “F1”, “OA” and “Kappa” remains unclean to reader without reading of page 7).
Author Response
We would like to thank for all comments. Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
I have no more questions.
