Using UAV Multispectral Remote Sensing with Appropriate Spatial Resolution and Machine Learning to Monitor Wheat Scab

Round 1

Reviewer 1 Report

 

1.       The manuscript title seems absurd. The authors are suggested to remove the word “research”.

2.       The manuscript needs thorough proof reading as it contains several long sentences.

3.       There are several typos throughout the manuscript like line 13 and 18” UAV multispectral” and “UVA multispectral” . Acronyms should be addressed properly as they are used in the manuscript. Line412: “vis” should be “VIs”. Correct Line 24 “period. in the multi-source…”. Line 241: “di” should be “DI”. Line 349: “BNPP” is “BPNN”. Line 412: “vis” is “VIs”.

4.       Its better if the authors only mentioned the best monitoring metrics only instead of mentioning 2^nd and 3^rd in abstract.

5.       Line83-91: The whole information can be phrased in two lines. Why the authors are using long sentence structure using commas.

6.       Line 83: “and spatial resolution directly affects image clarity and processing efficiency.” The line is mentioning a research gap. So better to write it as “how the spatial resolution directly effects the …….”

7.       Proper sentence structure is missing like line409-414 there is only one line. The sentence can be better interpreted if structured as 2 or 3 sentences.

8.       There’s no need to create another subsection data collection. It is understood that after overview of study area data will be collected. So make sec:2.2.1 as sec:2.2.

9.       No need of sec 2.3. Rename Sec2.3.1 as sec 2.3

10.   Line 153-156: Image resolution is expressed in cm. It should be cm/pixel or else. Moreover, the authors claim that as the height increases processing time increase and spatial resolution decreases. But the Table 1 referred is showing opposite of it. Please provide correct information or rewrite your claim.

11.   Sec: 2.3.4: The modeling method details failed to explain the overall methodology used by the authors. It would be great if the authors elaborate the steps using a flow diagram. What is SPXY sampling method?

12.   What is meant by section “2.3.5. accuracy verification”. Are these assessment metrics?

13.   Line222-223: “the average DI value of wheat scab” is mentioned twice.

14.   Line 232: is referring figure 3a or figure 3b. Please explain the analysis of results by referring the figure.

15.   Figure5: Spatial resolution is mentioned as cm pixel. Is it cm/pixel?

16.   Figure 8 Needs proper explanation in the text. The x-label is Flight altitude/m? why?

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

See attached PDF file

Comments for author File: Comments.pdf

Author Response

请参阅附件。

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript contains interesting data on UVA multispectral monitoring of wheat scab. The novelty of the research has been indicated in the Introduction. The results are original and may be useful in practice.

The manuscript is generally well organized. However, there are some detailed comments:

lines 104-105: Why were there 42 varieties and the authors wrote 3 (Sumai 3, Nanjing 8611, and 20fh0012)? Why were these varieties selected?

line 105: Were the two sampling areas different?

line 107: Have the authors tried to use naturally infected samples for the study?

line 180: It has been written "8 texture features in each band". How many bands were considered?

The results are quite clearly presented and described. However, maybe also other models/algorithms should be used to improve the accuracy.

The discussion of the results should be more detailed with more references.

The directions of further research should be more detailed.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

The authors evaluated scab monitoring with UAV multi-spectral system. I see the manuscript qualified for Agriculture MDPI. However, some suggestions are provided in the attached pdf file to improve the manuscript quality.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 5 Report

The experimental work presented in the Manuscript, entitled „ Research on wheat scab monitoring with UAV multi-spectral images based on suitable spatial resolution " is interesting research with some promising results. The article reports , the best  spatial  resolution  for  UVA  multispectral  monitoring  of  wheat  scab  was  determined  to  be 3.47cm, and then, based on the 3.47cm best resolution image, . In the multi-source data fusion model, the MLR algorithm has a poor monitoring effect, the BPNN algorithm is second, and the SVR algorithm has the best monitoring effect. The R² of the training set is 0.8318, the RMSE is 4.7738, and the RPD is 1.9203. The R² of the verification set is 0.8398, the RMSE is 3.1409, and the RPD is 2.9506, there are several shortcomings and modifications that should be included in order to enhance the final manuscript for the readers.

 

1-      The title of the manuscript should be modified. For example, the application of UAV multi-spectral images based on suitable spatial resolution and supported by machine learning models to monitor wheat scab

2-      Abstract

·         The sentences of the manuscript need to be arranged.

 For example , Line 15 to line 17: The vegetation index (VIs) and texture feature (TFs) with ……………and 7.67cm) should be moved after the first sentence at line 13. Because this is relate to M&M. As well as line 19-21:   VIs and TFs were used as input variables……………………………………..monitoring models should be moved after the above sentence.

·         Please present the best spectral index could be used to detect DI.

·         Line 24: in must be changed to ( In)

·         Line 26 :  The R ²  of the training set is 0.8318, the RMSE is 4.7738 , which model? Or across the three models

3-      Introduction

·         Line 41 to line 44, please add the citations?

·         Line 53 to line 54, please add the citations?

·         Line 56 : Baofeng et al. (2021) must be changed to Baofeng et al.  [6]. As well as the authors must be arranged the number of citations again.

·         Line 65 : Scab is more suitable for UAV remote sensing monitoring than other wheat 

diseases.  This is general sentence. The authors should support this sentence by the pervious researches.

·          The authors must present the previous work about the multiple linear regression (MLR), support vector machine regression   (SVR), and BP neural network (BPNN) in introduction section.

·          Please highlight in introduction, what is the novelty (originality) of the work? And what is new in your work that makes a difference in the body of knowledge? What has been done that goes beyond the existing research.

4-      Materials and methods

·         The materials and methods are well written

 

5-      Results and analysis

·         Line 240: The correlation should be changed to the relationship. The authors presented in table 3 the Coefficient of determinations is not correlation Coefficient.

·         Line 250 :  the title 3.2. Correlation analysis between vegetation index and DI value must be changed to 3.2 determination Coefficient of the relationships between selected vegetation indices and wheat scab disease index.

·         Line 262 :  Correlation analysis between texture features and DI value. Please the authors should present the data as determination coefficient or change the data to correlation Coefficient. As I see texture features and DI value is low related.

·         correlation coefficient is repeated several times under the section 3.3 please you could write R or r.

·         line 284 : Please increase the resolution of Figure 4 to be clear for the reader?

·         Figure 5 : please change  the title of the y axis to Coefficient of determination

·         Table 3 and 4: please present the value of R² as two digital numbers. Also RMSE   and RPD   in table 4.

6-      Discussion

·         Figure 8. Correlation between flight altitude images and DI, should be presented in Results and analysis

·         In general the discussion is poor discussed and it should be present as separate section as Results and analysis to cover all present data.

·         Please, write the practical applications of your work in a separate section, before the conclusions and provide your good perspectives.

7-      Conclusion

·         Please write about the limitation of this work under the section of conclusion

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

I approve the second version of the amnuscript

Author Response

Thank you for your valuable comments, which have greatly benefited us.

Reviewer 3 Report

The authors responded to all comments. However, the responses are not satisfactory. Furthermore, the authors did not make significant changes to the manuscript in line with the comments. My doubts were not properly addressed in the text of the manuscript.
I suggest the authors reconsider all previous comments.

Author Response

对于误解您的宝贵建议，我们深表歉意。现在我们详细修改了论文。请参阅附件。

Author Response File: Author Response.pdf

Reviewer 5 Report

1-The authors did not take the discussion part as carefully as they should be.
There is severe poverty in the discussion section. Authors should divide the discussion parts as found in the presentation of results.
Another alternative is for the authors to combine the results and discussion in one section.
2- The authors should support the discussion section with several previous studies.

3- Figure 8. should be presented in Results and analysis section.

 

Author Response

We are very sorry that the discussion part was not properly revised in the last revision. Over time, we have added a lot to the discussion.

Response 1-2: 

line 397-421, 433-438,446-472: The discussion content of our paper is indeed a little barren. So we added a lot of supplementary explanations to the discussion part.

 

Response 3:

We have presented Figure 8 in the discussion and analysis section.