Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

I would like to share my comments and suggestions on 'Generation of Ground Control Points using Binocular Camera' with the author/authors. This paper presents a method to generate virtual GCPS using two types of images acquired from binocular and monocular cameras installed on a drone. Literature research, an algorithm explanation, experimental results, and their discussion are all included in the paper. The paper is a valuable resource for utilizing its ideas and results in related areas.

However, research is deficient in the literature that is directly linked to VGCPs.  Therefore, the authors need to conduct a more thorough literature review regarding the creation and practical use of VGCPs. Furthermore, this paper appears to be a report on an experiment or a guide for program tools. These need to be improved before being published. Attached on separate pages are some revision recommendations that I provide to help with this.

 

Best regards.

 

Reviewer

 Additional comments:

Title The term “Virtual” should be included in the present title as follows “ ...Virtual Ground Control Points.., to distinguish from real GCPs.   Line 47 : There is a discrepancy between the designated references [20,21] The two references mentioned in the text are not related to the accuracy guidelines for drone survey results. Make sure the information is accurate and make the necessary substitutions.   Line 53: Figure 1 1)      Take away the question marks that are upside down in the parallelogram. 2)      Take a look at the logic of the flow chart. If reconstruction doesn't have GCPS and reference points, we'll move to the DEM as the next step. The creation of a DEM without GCPs is a logical contradiction.   Line 82: Figure 2 1)      According to Line 42, the open drop map only processes point clouds acquired with a monocular camera. Where is the pre-processing of point clouds from a binocular camera done?   2)      EXIFF in a square box should be rewritten as EXIF.     Different formatting between monocular and binocular images. In general, the amount of data does not have a significant effect on the format. Why were monocular and binocular images formatted differently? Apart from the quantity of data, was there any other specific reason for doing this?   Line 100: 2.3... the generation of VGCPs. 1)      It is not clear how to generate VGCPs using images from monocular and binocular cameras. Especially, regarding the part mentioned in 1.A theoretical supplementary explanation is necessary. 2)      Is it necessary to have both monocular and binocular cameras in order to obtain VGCPs?  If so, this would be very uneconomical since we would have to take the same target twice.   3)      It is not practical to fly at such a low altitude of 5 meters or 10 meters.   Line 144: Title of Figure 5 …binocular text… is might be correct   Line 150: missing unit. … has a focal length of 2.2?   Line 264: Figure 11 Why did you show only one image in Figure 11, even though it was acquired by a binocular camera? The same applies to Figure 13.   Line 208: Move the phrase It is necessary to move this phrase before line 200 in the context flow.   Line 214-215: absolute accuracy

1)      What does 'W/o' stand for?

2)      How did you get the result of 4.5% in line 255? The reason why absolute accuracy is low is not appropriate because the house is located on the road.

  Line 344: Abbreviations It's unnecessary because it's already well quoted in the text.   Line 348: Figures in Appendix A What are differences between three attached pictures, and what do you want to present with these figures? - The End -

–The end- 

Comments on the Quality of English Language

What does it mean in a second flight in line 107?

Author Response

Thank you very much for your feedback. Please see the attachment. Please accept my apologies for the error in the document I previously sent you. I have attached the correct version.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript is well written and the experimental data is presented very clearly, but I have a question regarding the virtual control points that I did not clarify. Comparing Figure 6 and section 2.3, step 8 ,line 122, corresponds to "Detect dependencies K-DTREE" in Figure 6, as well as subsequent "Deleting unreprojected points on the monocular lens" and "Deleting unreprojected points on the left binocular lens", with the aim of obtaining the corresponding points between the monocular camera and the binocular camera. Then, as described in "Select the center point that has the most occurrences in other images", a point near the center of the monocular image is obtained as the initial point (step 11, line 130),  using this point and its corresponding point in the binocular camera as the coordinate origin, calculate the coordinate difference between the other corresponding points within a 500 pixel range around it, and use its mean as the initial point coordinate shifting amount (step 13, line 133). If the calculation process of virtual control points is as I understand it to be, then binocular cameras play a role in controlling the size of objects, which is referred to as "relative accuracy" in the article. So how does binocular cameras improve absolute accuracy?

In addition, there are some details that need further editing.

1. Table 3 is missing units.

2. Appendix B is likely an editing error.

Author Response

Thank you very much for your feedback. Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Sir/Madam

 I would like to thank the Authors for their quick correction by reflecting most of my requests. However, I expect you to reflect my recommendations attached on a separate page.

 

Best regards.

 

Reviewer

 Additional comments:

Line 170, 172: Figure 4 and Figure 5 Re-proyection should be replaced by Re-projection in both Figures.  
Line 180: Delete – symbole in 6. Line 186: Figure 6. 1)      As shown in Figures 10 and 12, the radiometric characteristics of the left and right images of binocular camera are quite different. Therefore, the results can be also very different depending on which image, left or right, is used. 2)      Take away the question marks that are upside down in the parallelogram.   Line 253: Figure 14 and Line 253 I think, Figure 14 is not an appropriate example of comparing relative and absolute accuracy. Nevertheless, if you really want to distinguish between both accuracies, you may need to modify the phrase as follows: Figure 14 illustrates one house with a good relative accuracy, because the dimensions of a house can be measured accurately in an arbitrary coordinate system. However, the absolute accuracy is poor or not known, because the house is not represented in the correct geographical reference coordinate system.     

Line 261: What exactly does the sentence below mean?   Readers cannot easily understand it, so please you had better rephrase it. GCPs provide the actual value, which is not considered while improving the accuracy of the DEM.   Line 292: Units are missing in Table 1, Table 2, and Table 4

The unit m should be included at the end of three titles.

  Table 5 High in Table 5 should be represented as Height.

Reviewer

 –The end- 

Comments on the Quality of English Language Line 253:
Figure 14 illustrates one house with a good relative accuracy,
because the dimensions of a house can be measured accurately in an arbitrary coordinate system.
However, the absolute accuracy is poor or not known,
because the house is not represented in the correct geographical reference coordinate system.

Author Response

Thank you for your feedback. Please see the attachment. 

Author Response File: Author Response.pdf