The Development of a Defect Detection Model from the High-Resolution Images of a Sugarcane Plantation Using an Unmanned Aerial Vehicle
Round 1
Reviewer 1 Report
In the Introduction, the contribution of the paper should be presented in detail. The analysis of current literetures, e.g. [10]-[16], and the comparasion with the proposed approach should be discussed.
The main contribution of this work is to developed a framework combining many existing algorithms and provide a potential way for detecting defective areas in sugarcane plantation images. Why these algorithms are utilized instead of others? It is better to present some explanation or comparing experiments.
Author Response
Dear Reviewer 1
We really appreciate your comments. We truly believe that your comments can improve our manuscript to be more interesting and more understanding for readers on the introduction and the discussion parts. My responses to each of your comments can be seen in the attachment.
Kind regards,
bhoomin Tan-ut , Panomkhawn Riyamongkol
Authors
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The use of UAVs to study agricultural areas is of increasing interest. Despite the value in presenting research of this kind, the presentation would significantly benefit from some heavy editing. A couple of points in particular.
- The use of the word "defective" in the title and throughout the paper is awkward and misleading. Referring to a "defective detection model" make it sound like the model itself is defective. I think more effective language would be a "defect detection model." Similar change, from an adjectival to a noun form, throughout the paper, is important. Also, for example, in Figure 1, "Defective areas" should be "Defect areas".
- On page 18 the paper refers to "sample p-value" in Table 3. This was confusing to me. Usually a p-value refers to the a probability associated with a hypothesis test, and is a number in [0,1]. The p values here seem to be the values of p in equations (1) and (2). The language will need to be changed to clarify this.
The results were interesting, so working through the language and minor errors is encouraged.
Author Response
Dear Reviewer 2,
We would like to give our gratitude to your comments. From your comments, we can edit some confusing points and words so that readers will not be confused. Our manuscript cannot be better without your suggestion. My responses to each of your comments can be seen in the attachment.
Kind regards,
bhoomin Tan-ut , Panomkhawn Riyamongkol
Authors
Author Response File: Author Response.pdf
Reviewer 3 Report
Some comments for authors:
Line 77. The authors should put the resolution in pixels per cm2, because the term "Ultra-High-Definition quality" is very subjective and can change across time.
Line 117. The authors repeat unnecessarily the information that trees cause shadows.
Line 120. The authors should put a reference about LAB.
Line 126. The phrase "the algorithm is shown as a pseudo-code 1." should be "the algorithm is shown as a pseudo-code in Algorithm 1.", and the term "Shadow Detection" is wrongly typed.
Line 177. The authors should mention table 4 to facilitate readers' understanding of the datasets and their sizes.
Line 184, "Step 3: The statistical values such as mean and standard deviation are calculated from step 3". It probably should be from "step 2", and the authors should explain why they did not use other statistical values like variance, kurtosis, skewness, etc.
Line 189. The authors could explain how the threshold of 0.05 was defined.
Figure 9 could have three images original image, the result of specialists and shadow detection because, in the result of specialists, the grids with thick squares hide the shadow areas.
Line 365. Authors must improve the writing of the paragraph: "Therefore, the developed program must be used by avoiding trees because trees make a high precision in defective analysis."
Author Response
Dear Reviewer 3
We really appreciate your suggestions. We believe that our manuscript cannot be better without your comments. The responses according to your comments have been attached in the attachment.
Kind regards,
bhoomin Tan-ut , Panomkhawn Riyamongkol
Authors
Author Response File: Author Response.pdf
