Early-Stage Pine Wilt Disease Detection via Multi-Feature Fusion in UAV Imagery

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1.         L109-112: What is the specific impact of the expression of positional information on the detection of early-stage PWD, and how is positional information extracted and processed?

2.         L117-137: There is a lack of information on the slope and relief of the terrain in the study area. The UAV is flown from 100 to 240 meters, what was the basis for the change in flight altitude and whether it was affected by atmospheric conditions. What is the spatial resolution of the images.

3.         L141-156: Has the dataset production process considered the percentage of early-stage PWD in the images so that some data filtering and cleaning can be done.

4.         L172-181: The two parts of the prediction results (frequency domain feature prediction and combined feature prediction) are specifically how they are combined to arrive at the final prediction results.

5.         L190: Why crop a 5472 × 3648 pixels image to 800 × 800 pixels. Most classification datasets represented by ImageNet have images with an aspect of around 300 pixels. And most neural networks have an input pixels of 224 × 224.

6.         L407-421: In this paper, you have carried out tests of the YOLO series models, and whether other neural network models have been used for testing, such as the Transformer model. the Transformer model has performed well in the field of computer vision, and with its network architecture based on the attentional mechanism, it may be well suited for extracting the features of early-stage PWD as well.

7.         Whether the model is still valid for early-stage PWD detection in different seasons, growth cycles, regions, and terrain conditions.

Comments on the Quality of English Language

English expression needs to be improved.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Find Attached a Word file with text editing suggestions to improve readability

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Despite the file attached, it is advisable to perform an English language review to improve readability.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This is a well-written manuscript with a clear storyline, and all the methods and results are presented clearly. I only have the following minor comments:

Lines 15-21: I think this paragraph needs to be modified, as the abstract in its current version does not convey much about the importance and implications of the utilized approach in detecting PWD. The authors used the last sentence in their abstract to simply highlight the overall accuracy. I recommend rewriting the abstract.

If you have higher quality photos for the PWD stages, please update Figure 1.

Lines 292-294: Move this sentence to the methodology section.

In the discussion section, add the limitations of your study. We understand that using UAV is a promising approach as it can provide high spatial and spectral data; however, we also acknowledge that this approach is not applicable when dealing with a large area, or utilizing UAVs in some areas requires permission or is restricted. Please consider including a paragraph about this issue.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf