Three-Dimensional (3D) Parametric Modeling and Organization for Web-Based Visualization of City-Scale Pipe Network
Round 1
Reviewer 1 Report
The paper is well structured.
Style of subtitles are not in the MDPI rules. You must fix that.
Author Response
The paper is well structured.
Style of subtitles are not in the MDPI rules. You must fix that.
Reply: Thank you. The manuscript has been modified in both content and format.
Reviewer 2 Report
It seems that the paper presents a project from where the reader may extract valuable information, but at the end, I could not extract that information, since paper is bad written and is really hard to follow.
There is not an outline of the process, it would be helpful that the authors first introduce the steps of the process, the inputs, and the outputs in each step, making always clear the terminology. Some technical terms and processes are not introduced or defined, confusing the reader.
In general authors do not try to explain didactically the process. Complex procedures are solved with simply a sentence.
To be accepted, the authors must carry out a substantial rewriting of the article, trying to explain everything with more detail and didactically.
More in detail
Please, clarify the meaning of the terminology of equation (1). It is not clear what is SP1, SP2, are the distances from S to P1 (and P2)? Later, one realizes that by SP1, the authors means THE LINE connecting S and P1, but this must be introduced first to make this clear.
In addition, I guess that |SP1| is the length of the line, but what does it mean |SP1| with an arrow on it?
Line 128, it is not clear to me how S1 can be computed using SP1 and SP2 but without s (and without \theta), moreover, being s a parameter of the system. I am confused.
What is P4 (xp4,yp4,zp4) ? in equation (2)
And P3 in equation (3)?
Line 154, please, the paper must be self-contained. Please, define all terms in the paper, in this case “dynamic texture mapping implementation”.
Line 157, the sentence “the texture is assigned by a repeat” says practically nothing. And the rest “Here (u,v) represents…” doesn't say much either.
At the end of section 2.2.2, although authors said that in that section the texture mapping was presented, it was not presented, it just has two vague paragraphs.
Section 2.3 “parametric pipe-point modeling”, what is this? What is the “pipe-point prototype” and what is the ”pipe segment prototype”? If the authors do not make clear what is the meaning of the terminology, it is impossible to follow the paper.
Section 2.3.3, I do not know what is MmodelMatrix, Mscaling, Mrotation, Mtranslation, and how these apply to a flange.
Lines 211-212, I think: 3D Tiles are an open specification-> 3D Tiles is an open specification (IT IS an specification).
Line 214: 3D Tiles are built-> 3D Tiles is built.
Line 218: “pipelines should be partitioned at the pipe point”, in English, if you use the definite article (the), it should be clear which (one) pipe point the writer is referring to. Therefore, I think that the sentence should be “pipelines should be partitioned at pipe points”.
Line 219: It suddenly appears out of nowhere a quadtree, for what?
Line 220: This sentence “To achieve this goal, we improved upon the common quadtree partition rules, which are not strictly divided according to the same size, forming a loosely regulated quadtree organizational structure” does not make any sense. The partition rules of the quadtree are strict, I guess that you modified that, but the sentence does not say that, may be “ To achieve this goal, we improved upon the common quadtree partition rules, in such a way that quadboxes (areas or regions) at the same level of the tree now can have different sizes” … if I understand correctly.
This is a good example of the lack of precision of the authors. In Line 222, “The pipeline model can be loaded into blocks without partitioning the segments into different blocks using this method”, what this means? What is a block? I guess that block is a region, also known as quadbox, also known as areas. But a block in a quadtree is ambiguous, since it may refer the disk blocks that allocate the quadtree. In addition, the sentence is clearly confusing. I am an expert in quadtrees, and I can imagine what you are saying, but a normal reader will have problems. This subject must be clearly explained.
Also, if you need flexibility to avoid cutting segments, why not use an R-tree?, where regions (blocks?) are data driven.
Section 2.4.2 and figure 9 are useless, I did not understand nothing, unless you give more details. The only thing that I could figure out is that, if you need a spatial index (although I do not know for what) where regions (blocks?) are data driven, that is an R-tree.
Author Response
Thank you for your comments and suggestions. Response is uploaded. Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 3 Report
Cessium has recently introduced an improvement that enables visualizing underground data: https://cesium.com/blog/2020/06/16/visualizing-underground/
This study is one of the first case studies that focus on developing a pipeline data structure that can be digested into the Cesium as a 3D Tile. 3D visualization of pipelines by itself is not something new, but the use of 3D tiles can make it quite fast and efficient to render. By reading your paper, I should be able to replicate your developed approach to preparing a 3D tile for rendering in Cesium. It is possible if you can make a Github page for this research, share the code, and make a clean and comprehensive documentation of your step by step process for creating a web-based visualization of a pipeline network. It is great to see how your program can be used to quickly prepare 3D tiles from raw pipeline data.
In the introduction and method sections of the paper, you have discussed the advantage of the proposed approach and the problems of the previous studies. In the experiment section, it is great to compare the results of your approach with some of the other common methods/data structures to prove the efficiency of your proposed method.
The authors have done a great job in preparing high quality and informative graphics. However, the last two figures in the result section (Fig 12 and 13) can be improved and further explained. Providing a link to your web map as a live demo of your work can significantly helps to better experiment the results of your work, if possible.
Author Response
Point 1: It is possible if you can make a GitHub page for this research, share the code, and make a clean and comprehensive documentation of your step by step process for creating a web-based visualization of a pipeline network. It is great to see how your program can be used to quickly prepare 3D tiles from raw pipeline data.
Reply: First of all, thank you for reviewing the article. The software is still being improved, we need more time to strip the part of the pipeline processing code from the entire project. With the team’s approval, we will make a clean documents and code for creating a web-based visualization of pipeline network. After these are completed, we will make it open source on GitHub.
Point 2: In the introduction and method sections of the paper, you have discussed the advantage of the proposed approach and the problems of the previous studies. In the experiment section, it is great to compare the results of your approach with some of the other common methods/data structures to prove the efficiency of your proposed method.
Reply: It is difficult to know the specific implementation of other method, no direct comparison has been made. There are two points to the efficiency of this method, one is the efficiency of data processing, and the other is the efficiency of rendering. It takes about 400s to complete the process for city-scale pipeline data (mentioned in the article). As it is processed in advance, this part can fully meet the requirements of applications. In terms of rendering efficiency, we improved the quadtree index and organized the model into 3DTiles format, the average rendering FPS increased from 20 to 60 on the same computer.
Point 3: The authors have done a great job in preparing high quality and informative graphics. However, the last two figures in the result section (Fig 12 and 13) can be improved and further explained. Providing a link to your web map as a live demo of your work can significantly help to better experiment the results of your work, if possible.
Reply: Thank you. The Figure 12 and 13 are improved and further explained.
Due to data security issues, we provide a short online video to demonstrate the pipe visualization effects. The video address is
http://106.54.85.84:8080/proj/3D%20Pipeline/3D%20Pipeline.html .
Round 2
Reviewer 2 Report
Authors have addressed all my concerns.
Author Response
Thank you for reviewing the revised manuscript again.
Reviewer 3 Report
Thanks for your responses and the revisions. I have only one more comment:
Figure 13 should be further explained in the caption and if possible the Chinese language in the image should be translated.
Author Response
Thanks for your constructive suggestion. Figure 13 is updated and further explained, the upper part of figure shows the visualization of the cross-section in 3D scene, and the lower part is the result of the cross-section analysis. The Chinese language in the figure is translated into English.
