Change Detection in Urban Point Clouds: An Experimental Comparison with Simulated 3D Datasets†
Round 1
Reviewer 1 Report
Dear Authors
The presented work seems to be interesting but needs further corrections and considerations to become an acceptable paper on this journal.
1- Abstract does not present important points. It should be between 250 to 300 words and concisely mention the problems of previous works and novelties in this paper.
2- Similarly, the introduction has very poor structure and lack of literature review. Usually in the chapter of introduction the background and needs of study of change detection and damage assessment methods have to be highlighted and prepare readers to go further. Then your second chapter should be literature review where you present an overview on the previous works and the main problem statements of work and how it can be improved or overcome on it.
3- Please provide more information about the selected location and data repository and how they have been collected.
4- Results and discussion are not properly organized and it has to show the significant achievements of the proposed method and discuss each table and figure properly and in detail.
5- As you have lots of abbreviations, It is good that you have provided a table of abbreviations according to MDPI style.
6-It could be helpful if you do a comparison between your proposed method and some of the available or common other methods to show the efficiency of it.
7- It would be great if you provide a general framework or flowchart that how others can implement or use your proposed method for their assessment purposes.
8- In general, your conclusion is good and you can discuss a bit again about the achievements and novelty of your proposed method.
9- In total, the main problem of you paper is the lack of literature review and you can present some new developed methods for change detection, application of feed forward training and ML techniques in vulnerability and damage assessment of buildings, and infrastructures to attract the attention of readers and show a wide view of your works. Below are some of the recent works, where I found them new and useful to add and make your paper much more interesting:
-A Machine Learning Framework for Assessing Seismic Hazard Safety of Reinforced Concrete Buildings
-Evaluation of Change Detection Techniques using Very High Resolution Optical Satellite Imagery
-Application of open tools and datasets to probabilistic modeling of road traffic disruptions due to earthquake damage
-Post-earthquake road damage assessment using region-based algorithms from high-resolution satellite images
Author Response
Dear Reviewer 1,
Please see the attachment.
Best regards,
Author Response File: 
Author Response.pdf
Reviewer 2 Report
General comments
The study is interesting and potentially useful to the practitioners. In the results section, a critical comparison with the literature has been done. The discussion provided is sufficient for this publication.
The article is hard to follow due to the language use. The English should be improved. Therefore, I think this paper is not (yet) in the state for a journal publication.
Specific remarks:
RELATED WORK
Lines 80-82: These sentences do not make sense.
Line 87 (…detect changes into 3D data…’): Which 3d data? Urban point clouds, building, transport infrastructure…?
Line 87 (…at both dates…): Try to rephrase it.
Line 96: each class.
Lines 113-115: Give more references considering your argument.
Lines 116-132: This paragraph is mostly based on one reference ([17]). More should be added in this context.
Line 144: each point cloud.
Lines 159-163: More examples about the use of machine learning in this field should be given.
MATERIALS AND METHODS
A figure with the general workflow of the paper should be added, showing how the comparison between the different SoA methods has been done.
Simulation of changes in urban point clouds
Line 191: How was that simulator developed? With which tools?
Line 195: How is that extraction made? Do you apply already developed algorithms or are they new?
Line 203: Which is the workflow followed for the labelling processes? Which is the approach for the evaluation of changes?
Line 203: …between the two dates. Which two dates? Try to rephrase it, as said before.
Experimental protocol
A description of the methods applied should be given. A figure with the workflow followed could be very helpful, depending on if you make the general one recommended for the ‘materials and methods’ section.
RESULTS
This section reads more clearly than the rest of the article. Please, improve the wording in said sections.
Line 377: different levels of what?
Author Response
Dear Reviewer 2,
Please see the attachment.
Best regards,
Author Response File: Author Response.pdf
Reviewer 3 Report
1. About noise simulation. During real Lidar scaning, not only Gaussian noise exists. Please simulate more noise types in your analysis. For example, for different weather and temperature, the LiDAR noise is different.
2. Currently, the valid range of Lidar is limited to 70 meters. In near area, the scanning density is high and accurate. In far area, the density is low and rough. Did you consider such situation?
3. Using 3D model to simulate 3D LiDAR scanning results, did you consider that if the LiDAR scan on a surface of two sequantial frames, the two scanned lines on the same surface are different. Your testing algorithms may estimate such surface is have changed between two frames.
4. The paper lacks of novel contribution. The applied methods are typical, and not discribed in detail.
5. In segmenation process, if the 3D points are mapped onto 2D image, the segmentation of covering objects can't perform well.
Author Response
Dear Reviewer 3,
Please see the attachment.
Best regards,
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Dear Authors
The paper shows significant changes.
Reviewer 3 Report
The paper has been well revised based on the reviewers' opinions. All the problems were replied and described well in the revised version. I agree to accept this paper to be published.
