Detection and Location of Dead Trees with Pine Wilt Disease Based on Deep Learning and UAV Remote Sensing
Round 1
Reviewer 1 Report
This article introduced an approach to detect Pine Wilt Disease using deep learning technology and UAV. The authors did some improvement over the common faster-RCNN, such as using ResNet101 to replace VGG16, optimised anchor sizes and some modification of neural network. The methodology and results are convincing. The article is well written.
Two suggestions:
- Figure 3 is missing.
- Could the proposed entire detailed neural network be presented in the article such that interesting audience can reproduce or reuse the method in the article?
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The objectives of the presented work was to study the detection and orientation of pine wilt disease dead tree based on deep learning and UAV remote sensing. The study provides some interesting data for possible publication. However, the structure of the study needs to improve and several parts of the ms is necessary to revise before publication. Here I provide indicative examples for the revision.
Comments:
Abstract
L7-14: Give also full name of UAV, RCNN, RPN, VGG16.
Introduction
In this section, several abbreviations are used without giving the their full names.
L23: Give the latin name of nematodes causing PWD.
L28: A space is missing between ‘bacteria’ and ‘(Ponpandian…’. There are many places in the mes where these types of mistakes are exist (L29, L36, L43, L44, L60, L61. L127 … so on) . Please check and correct throughout the text.
L29: Delete R. and studied (not Studied)
L31: A space is missing ‘al.’ and ‘(2015)’.
L32: analysed (not Analyzed)
L32: Pinus massoniana – should be in italic
L40: pine (not Pine)
L42: machine (not Machine)
L48: Xiaohang Huang et al. (2018)… The same type of mistake in L55, L62. L69, L74.
L83: The clearly stated objectives are missing. Give the objectives here.
Materials and Methods
L94: trees (not Trees)
L96: There is no indication in the photos of figure 1 for a, b, and c
L86: What pine tree species were in the observed field?
L114: map (not Map)
L118: improvement (not Improvement)
L127: Girshick (not Girshick R)
L132: Delete the title of Figure 3. You do not need a figure for an equation. Equation belongs to the text. Change further figure numbering accordingly.
L144: Nine (not nine). More info is needed to include in the title of Figure 4.
L179: ()?? Explain
Results
L205: Results (not Experimental Results)
L224: Comparison (not Comparison). The title is not informative. More info is needed to include in the title of Table 1.
L245: PR – give in full. The title is not informative. More info is needed to include in the title of Figure 6.
L249: The title is not informative. More info is needed to include in the title of Figure 7.
L249-L260: You should merge Figures 7-9 into one Figure.
Discussion and Conclusion
L262: I think you should divide Discussion and Conclusion. First Discussion then Conclusion.
L263-274: The whole section should be moved after Discussion. You need to give clear point by point conclusions.
L275-L291: Discussion is short and do not cite any previous results. This section is necessary to fully rewrite. In the present form, it is not a Discussion.
References
This section does not follow the mdpi requirements.
Several other formatting points are also missing.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
Most of the required minor changes were made.
Three essential aspects did not changed.
1) The Discussion Section is still not a discussion. There is no citation in this section. Authors should compare their results with previous studies and cite many previous literatures to support their results. It should be at least two pages long.
2) Conclusion is too long. Should be half of the length.
3) References are not consistent, did not follow unified format according to mdpi requirements.
Author Response
Please see the attachment
Author Response File: Author Response.docx
Round 3
Reviewer 2 Report
The manuscript is improved in an acceptable level.
