Multi-LiDAR Mapping for Scene Segmentation in Indoor Environments for Mobile Robots

Round 1

Reviewer 1 Report

1. The readability is poor, so it is difficult to understand. Overall, it is necessary to describe the paper concisely and clearly.
2. There are double spaces everywhere, so they need to be corrected.
3. To verify the performance of the proposed method, performance comparison in case of 2D, 3D, and fusion of 2D and 3D sensors are necessary. In addition, it is necessary to select a more complex scenario and to show the results of applying various scenarios.
4. The description of the proposed algorithm is insufficient. For example, it would be good to add a schematic diagram of the entire proposed algorithm.
5. Most of figures too small to difficult to see and distinguish. In addition, the results analysis for the figures are insufficient.
6. It would be good to add some recent references related to Sensors.

Author Response

Thanks to the reviewer comments and guidelines to improve the paper structure and quality.

 

The readability is poor, so it is difficult to understand. Overall, it is necessary to describe the paper concisely and clearly.

 

Authors appreciate the reviewer suggestions. English have been reviewed and additional text have been added to improve the readability.

 

There are double spaces everywhere, so they need to be corrected.

 

Format have been reviewed to avoid double spacing.

 

To verify the performance of the proposed method, performance comparison in case of 2D, 3D, and fusion of 2D and 3D sensors are necessary. In addition, it is necessary to select a more complex scenario and to show the results of applying various scenarios.

 

Additional data and a resume table of the simulated experiment accuracy have been added to show the performance of the algorithm also have been added an explanation of the complexity of both simulated and real-world scenarios.

 

The description of the proposed algorithm is insufficient. For example, it would be good to add a schematic diagram of the entire proposed algorithm.

 

Two new flow charts and their descriptions have been added to help to understand the overall process and algorithms.

 

Most of figures too small to difficult to see and distinguish. In addition, the results analysis for the figures are insufficient.

 

Figures have been enlarged and analysis have been expanded.

 

It would be good to add some recent references related to Sensors.

 

Sensors references have been added.

Authors Thanks again to the reviewers, and hope the changes that have been made fit and improve the quality of our research.

Reviewer 2 Report

The manuscript presents a method that takes advantage of multi-Lidar and SLAM for room/scene segmentation in an indoor environment. The rationale behind the study is well articulated and the research method is clearly described. This reviewer enjoyed reading the manuscript and would like to support the publication of this work after addressing the following points.

• Consider putting the abbreviations at the beginning of the manuscript or within the content (where they first show up).
• Include a glimpse of the limitations/weakness of present work and near-future plans, what can people build upon your work.
• The comparison between the existing method vs the proposed one could be improved. For instance, while people in the field may readily understand the difference between figure 14.a and 14.b, it would be nice to point out (e.g., circle or highlight, etc.) the key differences.

Author Response

Thanks to the reviewer comments and guidelines to improve the paper structure and quality.

The manuscript presents a method that takes advantage of multi-Lidar and SLAM for room/scene segmentation in an indoor environment. The rationale behind the study is well articulated and the research method is clearly described. This reviewer enjoyed reading the manuscript and would like to support the publication of this work after addressing the following points.

The authors appreciate that this reviewer sees the potential of out work.

Consider putting the abbreviations at the beginning of the manuscript or within the content (where they first show up).

Text have been reviewed to acknowledge abbreviations order of appearance

Include a glimpse of the limitations/weakness of present work and near-future plans, what can people build upon your work.

Future works aims to optimize algorithms run-time.

The comparison between the existing method vs the proposed one could be improved. For instance, while people in the field may readily understand the difference between figure 14.a and 14.b, it would be nice to point out (e.g., circle or highlight, etc.) the key differences.

Figures have been enlarged and analysis have been expanded.

Authors Thanks again to the reviewers, and hope the changes that have been made fit and improve the quality of our research.

Reviewer 3 Report

The authors have proposed and implemented a multi-LiDAR based approach for indoor mapping. The proposed approach utilizes several techniques like Harmony Search and Simultaneous Localization and Matching. It also uses 2D and 3D sensors of the robot located at different heights.

The article is well-structured. The authors reviewed related research work from recent years (although the most recent work is from 2019). On this basis, they formulated a research problem and proposed a solution based on the existing algorithms, which, however, ensures the avoidance of various problems related to mapping the interior of buildings and rooms. The proposed solution was verified experimentally based on many scenarios, both virtual and in the real world.

The article is interesting, but the following issues should be addressed before publication:

1. The English language should be corrected as there are grammatical and stylistic errors.

2. Are there no metrics that would make it possible to compare the quality of solutions provided by individual algorithms? Currently, the results are presented in the form of maps that should be interpreted and compared qualitatively. However, there are definitely no metrics that would make it possible to conduct not only qualitative but also quantitative comparisons of the quality of solutions. If such metrics do not exist at present, then they could be proposed in future research works.

Author Response

Thanks to the reviewer comments and guidelines to improve the paper structure and quality.

The article is interesting, but the following issues should be addressed before publication.

The authors appreciate that this reviewer sees the potential of out work.

The English language should be corrected as there are grammatical and stylistic errors.

Authors appreciate the reviewer suggestions. English have been reviewed and additional text have been added to improve the readability.

Are there no metrics that would make it possible to compare the quality of solutions provided by individual algorithms? Currently, the results are presented in the form of maps that should be interpreted and compared qualitatively. However, there are definitely no metrics that would make it possible to conduct not only qualitative but also quantitative comparisons of the quality of solutions. If such metrics do not exist at present, then they could be proposed in future research works.

New tables and description have been added, showing the accuracy of the results for registration and the metrics for door segmentation.

Authors Thanks again to the reviewers, and hope the changes that have been made fit and improve the quality of our research.

Round 2

Reviewer 1 Report

The authors seem to have corrected all the comments well overall.