Spatiotemporal Assessment of Satellite Image Time Series for Land Cover Classification Using Deep Learning Techniques: A Case Study of Reunion Island, France
, Kandasamy Chandrasekaran 1, Andrzej Stateczny 2,*, Venkatesan Meenakshi Sundaram 3 and Prabhavathy Panneer 4
Round 1
Reviewer 1 Report
The article is interesting and makes an important contribution to the field, but requires an improvement of the presentation of methods and discussion of results before its publication. Detailed comments are provided for each issue.
1. Line 139 states that there are two research goals, but there are three goals explained in lines 140-149; please clarify.
2. Please provide in section 2, Materials and Study Area, arguments for the choice of your study area (e.g., has specific land use patterns etc.)
3. Please show the position of Reunion Island in Figure 1 in a broader context, showing other neighboring land areas with their names, using the map-in-map system.
4. In addition to the quantitative comparison of the spatiotemporal deep learning models presented in Tables 2 and 3, the manuscript would benefit upon a qualitative comparison, summarizing in words the advantages and disadvantages or limits of each model.
5. Please expand the discussions to reveal: (A) the inner validation of results, against the study goals presented in lines 140-149; (C) the external validation of results, against those of similar studies from other countries, identified in the literature; (D) the importance of results, meaning their contribution (conceptual or methodological) to the theoretical advancement of the field; (E) a summary of the study limitations and directions for overcoming them in the future research.
Author Response
Response to reviewer-1
- Reviewer comment: Line 139 states that there are two research goals, but there are three goals explained in lines 140-149; please clarify.
Answer: The following changes are incorporated at Line no.143 in the revised manuscript.
- Reviewer comment: Please provide in section 2, Materials and study Area, arguments for the choice of your study area (e.g., has specific land use patterns etc.)
Answer: The motivation behind choosing the study area was due to the variation in the land use patterns for effective classification of land covers being addressed by our proposed approach. changes are incorporated at Line no.198-229 in the revised manuscript.
- Reviewer comment: Please show the position of Reunion Island in Figure 1 in a broader context, showing other neighboring land areas with their names, using the map-in-map system.
Answer: The following changes are incorporated at Line no.231 in the revised manuscript.
- Reviewer comment: In addition to the quantitative comparison of the spatiotemporal deep learning models presented in Tables 2 and 3, the manuscript would benefit upon a qualitative comparison, summarizing in words the advantages and disadvantages or limits of each model.
Answer: The following changes are incorporated at Line no.459-483 in the revised manuscript.
- Reviewer comment: Section 3.2.3: Please expand the discussion to reveal: (A) the inner validation of results, against the study goals presented in lines 140-149; (C) the external validation of results, against those of similar studies from other countries, identified in the literature; (D) the importance of results, meaning their contribution (conceptual or methodological) to the theoretical advancement of the field; (E) a summary of the study limitations and directions for overcoming them in the future research.
Answer: The following changes are incorporated and the discussion section is expanded from Line No.507 in the revised manuscript.
Reviewer 2 Report
This study (Spatio-temporal Assessment of Satellite Image Time Series for Land Cover Classification using Deep Learning Techniques: A Case Study of Reunion Island, France.)
The paper is good presented, and the results are good explained. However, it is necessary to make some modifications in order to be accepted. Corrections are listed below:
1. Check line 21
2. L26-28: Check
3. What is RNN in line 30
4. Why “Spatio-temporal” in capital letters?
5. All keywords are similar to the article title and the authors should add different keywords so the paper can be widely searchable
6. Introduction section need to be organized in subtitles and more newly references should be added to the introduction
7. What is SITS in L42?
8. What is RS in line 62?
9. Citations style is wrong in the paper. Please revise it.
10. What is LCC in line 89?
11. Why “Other” and “Water” in capital letters in line 178?
12. Figure 1 should be modified. Add Lat and long, features of the study area, north arrow and legend
13. Align the text in figure 4
14. Change * to × in figure 5
15. Insert table 2 and 3 after line 314
16. Figure 6 and 7: please revise the alignment
17. Check the text format in the conclusion section
18. Please check the references as there are some references without Doi. Please check them carefully.
Author Response
Response to reviewer-2
- Reviewer comment: Check line 21
Answer: The following changes are incorporated at Line no.21 in the revised manuscript.
- Reviewer comment: L26-L28: Check.
Answer
The following changes are incorporated at Line no.26-28 in the revised manuscript.
- Reviewer comment: What is RNN in line 30
Answer:
(Recurrent Neural Network) The following changes are incorporated at Line no.31 in the revised manuscript.
- Reviewer comment: Why Spatio-temporal in capital letters.
Answer:
The following changes are incorporated at all positions in the revised manuscript.
- Reviewer comment: All keywords are similar to the article title and the authors should add different keywords so the paper can be widely searchable.
Answer:
The following changes are incorporated at Line no.37 in the revised manuscript.
- Reviewer comment: Introduction section need to be organized in subtitles and more newly references should be added to the introduction.
Answer:
The following changes are incorporated at Line no.70 and 107 in the revised manuscript.
- Reviewer comment: What is SITS in L42?
Answer:
Satellite Image time series (SITS). The following changes are incorporated at all locations in the revised manuscript.
- Reviewer comment: What is RS in line 62?
Answer:
(Remote sensing) The following changes are incorporated at Line no.63 in the revised manuscript.
- Reviewer comment: Citation style is wrong in this paper. Please revise it.
Answer:
Sir, the citation style is retrieved as per the template from the MDPI site. If not found proper the citations will be updated as per the journal guidelines as well as per advice from associate editor if not in proper form before the acceptance of the manuscript.
Round 2
Reviewer 2 Report
Accept in current format
