Long-Range 3D Reconstruction Based on Flexible Configuration Stereo Vision Using Multiple Aerial Robots

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper introduces a long-range 3D reconstruction method using multiple aerial robots. The proposed method, system design and results are well organized. Generally, this is a well-written manuscript with good results. There are only few suggestions.

 

1. Page 14, Figure 15, Depth estimation and their corresponding match percentage of area of interest are presented. In this figure, box charts of different baseline configurations are presented. As I see, the performance of different baselines varies.  For example, for POI B, as the vertical baseline increases, it seems the estimation accuracy is not simply increasing. It would be better analysis the reason.

2.  Since the observed scene have influence on the performance of proposed method. Please add some discussions regarding what kind of baseline configuration should be adopted when facing different scenes.

Author Response

Please refer to the attached PDF file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The work presented regarding the long-range 3D Reconstruction based on Flexible Configuration Stereo Vision using Multiple Aerial Robots certainly has the merit of improving the acquisition capabilities present today. The introduction appears concise and releva and the explanatory phase of the methodology used is well structured both in the description of its various phases and the parameters used. The results achieved appear very interesting, certainly capable with the continuation of research of expressing further and better performances in the future. The bibliographical references, although a little limited, are still relevant and the English is understandable. Publication in the current version is recommended.

Author Response

Please refer to the attached PDF file.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

An interesting article showing the results of the authors' research on improving 3D reconstruction using aerial robots or unmanned aerial vehicles (UAVs). In times of creation and use of augmented and virtual reality, this is undoubtedly a good direction of research. As you can see in the conclusions, there are still many problems to solve, but the authors' team will certainly gradually solve them.

 

Is Depth estimation of a distance up to 400 m with less than 10% error using only 10 m of active flight distance sufficient for applications in 3D reconstruction tasks? This gives an error of ±40 m, which is disqualifying for geodetic monitoring applications.

Components of figures 6,7,9,10,12,13 are too small and illegible.

What can actually be seen in Fig. 17 and 18. The point cloud shown seems to be something different than the photograph in Fig. 16 and 18. It is difficult to identify the objects presented in front of the point cloud.

Fig. 18 there is …resultant point cloud from point cloud fusion (left) – it should be (right)

Author Response

Please refer to the attached PDF file.

Author Response File: Author Response.pdf