An Adaptive Threshold Line Segment Feature Extraction Algorithm for Laser Radar Scanning Environments
Round 1
Reviewer 1 Report
Comments # 01: The introduction section is not clear. It is very broad and not at all comprehensively focused to relevance of the manuscript. It lists a lot of other people's research work, but it does not impress on readers the importance of this work, what is not solved in this field and why is it important. Therefore, as a reader, it is difficult to draw the conclusion from them as to why this study has been carried out.
Comments #02: The following latest studies are very relevant. The authors must read and provide complete information on this topic through including these studies. Intelligent computing technique for solving singular multi-pantograph delay differential equation, Neuron Analysis of the Two-Point Singular Boundary Value Problems Arising in the Thermal Explosion’s Theory, Artificial neural network scheme to solve the nonlinear influenza disease model, A Numerical Study of the Fractional Order Dynamical Nonlinear Susceptible Infected and Quarantine Differential Model Using the Stochastic Numerical Approach, A novel computing stochastic algorithm to solve the nonlinear singular periodic boundary value problems, Applications of neural networks for the novel designed of nonlinear fractional seventh order singular system, Computational intelligence approach using Levenberg–Marquardt backpropagation neural networks to solve the fourth‑order nonlinear system of Emden–Fowler model.
Comments #03: The research problem in the paper does not seem to be motivated by a clearly outlined research question and no physical insight is provided for this theoretical analysis.
Comments #04: Authors should not only state what is shown, but it should be made clear to readers why the figure has been included and what is of interest.
Comments # 05: A schematic figure representing the physical phenomenon of problem studied needs to be included.
Author Response
Dear reviewers:
We are very grateful to your comments for the manuscript. According with your advice, we tried our best to amend the relevant part and made some changes in the manuscript. These changes will not influence the content and framework of the paper. All of your questions were answered below. And here we list the changes and marked in yellow in revised paper.
We appreciate for Reviewers’ warm work earnestly, and hope that the correction will meet with approval. Should you have any questions, please contact us.
Once again, thank you very much for your comments and suggestions.
Yours Sincerely,
Lei Zhang
Comments # 01: The introduction section is not clear. It is very broad and not at all comprehensively focused to relevance of the manuscript. It lists a lot of other people's research work, but it does not impress on readers the importance of this work, what is not solved in this field and why is it important. Therefore, as a reader, it is difficult to draw the conclusion from them as to why this study has been carried out.
Author action:
Conceptualization, Yiting Liu, Lei Zhang, Kui Qian and Lianjie Sui; Methodology,Yiting Liu and Lei Zhang;Software,Lei Zhang and Kui Qian; validation,Yuhao Lu,Fufu Qian and Tingwu Yan; Writing—original draft,Yiting Liu and Lei Zhang; Writing – review and editing, Kui Qian,Hanqi Yu and Fangzheng Gao.
We study the overall design scheme of laser radar line segment feature extraction, and read the relevant literature. PDBS algorithm and IEPF algorithm in the existing literature give a fixed threshold to judge the feature points, which is easy to cause over-segmentation and under-segmentation. In this work, we propose an adaptive threshold method to determine the line segment features, which avoids the iterative calculation of IEPF, improves the accuracy of line segment feature extraction and reduces the calculation amount. We use SICK TIM571 laser radar in the real environment to verify the effectiveness of the algorithm. Other authors edit and modify the article.
Comments #02: The following latest studies are very relevant. The authors must read and provide complete information on this topic through including these studies. Intelligent computing technique for solving singular multi-pantograph delay differential equation, Neuron Analysis of the Two-Point Singular Boundary Value Problems Arising in the Thermal Explosion’s Theory, Artificial neural network scheme to solve the nonlinear influenza disease model, A Numerical Study of the Fractional Order Dynamical Nonlinear Susceptible Infected and Quarantine Differential Model Using the Stochastic Numerical Approach, A novel computing stochastic algorithm to solve the nonlinear singular periodic boundary value problems, Applications of neural networks for the novel designed of nonlinear fractional seventh order singular system, Computational intelligence approach using Levenberg–Marquardt backpropagation neural networks to solve the fourth‑order nonlinear system of Emden–Fowler model.
Author action:
Thank you for your suggestion. We have read the latest relevant research you have provided and included complete information on this topic in subsequent in-depth studies.
Comments #03: The research problem in the paper does not seem to be motivated by a clearly outlined research question and no physical insight is provided for this theoretical analysis.
Author action:
The line segment feature extraction of laser radar has an important influence on the accuracy of mapping and the positioning of robots in the environment. The PDBS algorithm and IEPF algorithm in the existing literature have given a fixed threshold to determine the feature points, which is easy to cause over-segmentation and under-segmentation. In this work, we propose an adaptive threshold method to determine the line segment features, which avoids the iterative calculation of IEPF, improves the accuracy of line segment feature extraction and reduces the calculation amount.
Comments #04: Authors should not only state what is shown, but it should be made clear to readers why the figure has been included and what is of interest.
Author action:
When the mobile robot navigates autonomously in an unknown environment, it needs accurate maps. At the same time, in order to construct accurate maps, the pose of the robot must be determined. The autonomous navigation algorithm based on laser radar is one of the research directions. Laser radar has the advantages of high ranging accuracy and long ranging distance. However, due to the small amount of data of laser radar and the influence of the noise of the sensor itself, it is easy to lead to low accuracy of the robot to build a map and large positioning error. In this paper, the point set in laser radar scanning environment is filtered, and the line segment feature extraction is completed by using adaptive threshold segmentation point set. This method ensures the real-time performance of mobile robot map construction and serves for subsequent positioning and navigation.
Comments # 05: A schematic figure representing the physical phenomenon of problem studied needs to be included.
Author action:
As shown in the figure, a laser radar frame data due to noise, there will be some noise, point set is not evenly distributed on the surface of the obstacle, need to be processed. In the process of line segment feature extraction, we need to extract the breakpoint and corner features, and segment the original point set. Most of the existing algorithms in the literature judge breakpoints and corners by a fixed threshold, which is easy to over-segmentation or under-segmentation.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Author presents the results of the analysis of an adaptive threshold line segment feature extraction algorithm for laser radar scanning environment. Article and the methodology is based on assumptions and on proposed scenarios. Some major errors need to be modified in paper.
Page n, Row n: At first, it should be denoted, that this paper was sent before the authors’ reviewing and reading. There are a lot of misunderstanding statements, sentences are raw translation and copy-pasted from some documents. The formatting of text is without any correction and proofreading (words without spaces; connected with signs or numbers; …). The paper is difficult to read and uses meaningless sentences (draft version?). The article deserves to be rejected (2 of 10 points), but I estimate that the authors can edit the article in a readable form and it will be suitable for publication.
Page 1, Row 15-27: Abstract must be extended. In abstract is one sentence in 5 rows – modify it.
Page n, Row n: Authors did not respect the meaning of variables, functions, units representation. In the whole text, equations, graphs, schemes the variables, functions, lower/upper indices … are written randomly. Please, use the proper denotation because it is unreadable, what author would like to present.
Page n, Row n: Use the proper sign for scalar multiplying (not vector multiplying).
Page n, Row n: Add units for variables (after expression definition).
Page 3, Row 13: What represents equation 1? What is a function? Which variables are multiplied and what are the units? In addition, the next expression are in the same manner and are meaningless.
Page n, Row n: Unify the equations size and size of expressions/equations in text, because they are copy-pasted from another document and they do not correspond to paper template.
Page 6, Row 206: Description of figure “This is a figure.” is not sufficient. What represents x and y-axes? What is legend and what is unit?
Page 6, Row 212: Why authors chose angular resolution of 0.33 degrees? Describe it.
Page 7. Row 223-230: One sentence per 8 rows. These long sentences are useless and should be divided into readable ones.
Page 7, Row 246-247: What type of symbols used authors? Is it apostrophe, comma, top line, or what? It is similar in expressions (12) and (14).
Page 7, Row 245: Point “O” is denoted in Fig. 5 “0”. Unify them.
Page 9, Row 282: Description of Fig. 6 is insufficient.
Page 10, Row 311: Where is a flow chart? What represents the cluster of words? It is not a flow chart. However, title of this part is not flowchart of “this paper”, but flowchart of algorithm or method. Revise the whole part of this section.
Page 15, Row 396: What represents the first sentence? Why authors did not read the text before submitting?
Page 17, Row n: The conclusion is just statement, without any discussion. Extend the conclusion.
Page 17, Row 44-449: Authors denoted a lot of funding grants, but the benefit of this paper does not correspond to any visible and relevant result.
Page 17-18, Row n: It is not possible to identify some references. Add additional information to references (authors, paper title, doi, volume, link, …) and align them according to template.
The topic of the paper is actual, but the paper implementation is horrible. Authors need to respond to all below mentioned comments, because this version is just draft paper version. The paper needs to be revised in general.
Author Response
Dear reviewer:
We are very grateful to your comments for the manuscript. According with your advice, we tried our best to amend the relevant part and made some changes in the manuscript. These changes will not influence the content and framework of the paper. All of your questions were answered below. And here we list the changes and marked in yellow in revised paper.
We appreciate for Reviewers’ warm work earnestly, and hope that the correction will meet with approval. Should you have any questions, please contact us.
Once again, thank you very much for your comments and suggestions.
Yours Sincerely,
Lei Zhang
Concern # 1:
Page n, Row n: At first, it should be denoted, that this paper was sent before the authors’ reviewing and reading. There are a lot of misunderstanding statements, sentences are raw translation and copy-pasted from some documents. The formatting of text is without any correction and proofreading (words without spaces; connected with signs or numbers; …). The paper is difficult to read and uses meaningless sentences (draft version?). The article deserves to be rejected (2 of 10 points), but I estimate that the authors can edit the article in a readable form and it will be suitable for publication.
Author action:
Thank you for giving us the opportunity to modify. We corrected some incorrect statements and formats, and the modified part was marked in yellow in pdf.
Concern # 2:
Page 1, Row 15-27: Abstract must be extended. In abstract is one sentence in 5 rows – modify it.
Author action:
Page 1, Row 15-22:We extend the summary by adding a background for laser radar line segment feature extraction and adjusting the contents of lines 18-22 :
At present, the feature extraction of environmental line-segment based on radar scanning data generally adopts the idea of recursion. However, the amount of calculation for recursion is large, and the threshold of extracted feature points needs to be set manually. Moreover, the fixed segmentation threshold will cause under-segmentation or over-segmentation.
Concern # 3:
Page n, Row n: Authors did not respect the meaning of variables, functions, units representation. In the whole text, equations, graphs, schemes the variables, functions, lower/upper indices … are written randomly. Please, use the proper denotation because it is unreadable, what author would like to present.
Author action:
We delete formula ( 1 ) ( 2 ) ( 3 ) in the original paper, which introduces the concept of Gaussian white noise and is not closely related to the content of the algorithm in this paper. The equations and variables in the text are checked and modified.
Concern # 4:
Page n, Row n: Use the proper sign for scalar multiplying (not vector multiplying).
Author action:
We deleted the original formula ( 1 ) ( 2 ) ( 3 ) and checked the multiplication symbol of the formula.
Concern # 5:
Page n, Row n: Add units for variables (after expression definition).
Author action:
The units of variables in this paper are mostly meters, but in the format of electronics, we do not see where the article adds units to the variables.
Concern # 6:
Page 3, Row 13: What represents equation 1? What is a function? Which variables are multiplied and what are the units? In addition, the next expression are in the same manner and are meaningless.
Author action:
Page 3, Row 15-22:We delete formula ( 1 ) ( 2 ) ( 3 ) in the original paper, which introduces the concept of Gaussian white noise and is not closely related to the content of the algorithm in this paper.
Concern # 7:
Page n, Row n: Unify the equations size and size of expressions/equations in text, because they are copy-pasted from another document and they do not correspond to paper template.
Author action:
We unify the formula and variable size in mathtype.
Concern # 8:
Page 6, Row 206: Description of figure “This is a figure.” is not sufficient. What represents x and y-axes? What is legend and what is unit?
Author action:
Page 5, Row 191-200:The figure is the characteristics of laser radar scanning to the laser point. The polar coordinate data collected by the laser radar are converted to the rectangular coordinate system. The measured distance is relative to the coordinates of the laser radar on the x-axis and y-axis. The x-axis and y-axis are the rectangular coordinate system, and the unit is m.
Concern # 9:
Page 6, Row 212: Why authors chose angular resolution of 0.33 degrees? Describe it.
Author action:
Page 6, Row 210-214:When the angle is too large, the laser points are dispersed, which may lead to the loss of key shape characteristics of the environment. Therefore, laser radar with small angular resolution should be selected. In order to meet the acquisition of indoor environmental characteristics, the angular resolution of laser radar used in this paper is 0.33°.
Concern # 10:
Page 7. Row 223-230: One sentence per 8 rows. These long sentences are useless and should be divided into readable ones.
Author action:
Page 7, Row 225-235:We simplified the long sentence.
Concern # 11:
Page 7, Row 246-247: What type of symbols used authors? Is it apostrophe, comma, top line, or what? It is similar in expressions (12) and (14).
Author action:
Page 7, Row 254:We changed formula ( 9 ) ( 11 ) and the symbols in the article to apostrophes.
Concern # 12:
Page 7, Row 245: Point “O” is denoted in Fig. 5 “0”. Unify them.
Author action:
Page 7, Row 248:The point O in the article and graphics are unified.
Concern # 13:
Page 9, Row 282: Description of Fig. 6 is insufficient.
Author action:
Page 9, Row 280-286: We give a supplementary description of Figure 6.
Concern # 14:
Page 10, Row 311: Where is a flow chart? What represents the cluster of words? It is not a flow chart. However, title of this part is not flowchart of “this paper”, but flowchart of algorithm or method. Revise the whole part of this section.
Author action:
Page 10, Row 317-323:We have modified this section, mainly a detailed description of the algorithm.
Concern # 15:
Page 15, Row 396: What represents the first sentence? Why authors did not read the text before submitting?
Author action:
Page 10, Row 441:We delete this sentence, thank you for pointing out the error.
Concern # 16:
Page 17, Row n: The conclusion is just statement, without any discussion. Extend the conclusion.
Author action:
Page 16, Row 489-499:We extended the conclusions and discussed them.
Concern # 17:
Page 17, Row 444-449: Authors denoted a lot of funding grants, but the benefit of this paper does not correspond to any visible and relevant result.
Author action:
These funded funds have topics and projects in the multi-sensor fusion localization of mobile robots. This paper studies the feature extraction of laser radar line segments to serve the subsequent multi-sensor fusion localization algorithm.
Concern # 18:
Page 17-18, Row n: It is not possible to identify some references. Add additional information to references (authors, paper title, doi, volume, link, …) and align them according to template.
Author action:
We corrected the information of references and aligned them with templates.
Author Response File: Author Response.pdf
Reviewer 3 Report
1. Line 15-18: I prefer to say it like this,
At present, the feature extraction of environmental line-segment based on radar scanning data generally adopts the idea of recursion. However, the amount of calculation for recursion is large, and the threshold of extracted feature points needs to be set manually. Moreover, the fixed segmentation threshold will cause under-segmentation or over-segmentation.
- Line 25: The algorithm -> The new algorithm
- Line 34, 35: This sentence is duplicated with the first sentence.
- Line 60: What is KF? Please provide its full name for the first-time use
- Line 69: line segment tracking -> LT
- Figure 1: The authors need to include the definition of all symbols, i.e., theta, .., of the figure in the caption.
- Line 122: cannot find alpha in Fig. 1
- Line 323: delete “for”
- Line 328: delete “there is”
- Line 373: Same as line 328, delete “there is”
11: Line 388: fit -> fits
12: Line 396: Not sure what this sentence is used for.
13: Section 5: The author should include all important conclusions in this section and the abstract. This section and abstract seem to emphasize different points. For example, the abstract highlights the similarity of the new method compared to the other methods, while the conclusion mostly focuses on the achieved accuracy.
Author Response
Dear reviewer:
We are very grateful to your comments for the manuscript. According with your advice, we tried our best to amend the relevant part and made some changes in the manuscript. These changes will not influence the content and framework of the paper. All of your questions were answered below. And here we list the changes and marked in yellow in revised paper.
We appreciate for Reviewers’ warm work earnestly, and hope that the correction will meet with approval. Should you have any questions, please contact us.
Once again, thank you very much for your comments and suggestions.
Yours Sincerely,
Lei Zhang
Concern # 1:
Line 15-18: I prefer to say it like this,
At present, the feature extraction of environmental line-segment based on radar scanning data generally adopts the idea of recursion. However, the amount of calculation for recursion is large, and the threshold of extracted feature points needs to be set manually. Moreover, the fixed segmentation threshold will cause under-segmentation or over-segmentation.
Author action:
Line 18-22:Thank you for your suggestion. We modified it.
Concern # 2:
Line 25: The algorithm -> The new algorithm
Author action:
Line 29:Thank you for your suggestion. We modified it.
Concern # 3:
Line 34, 35: This sentence is duplicated with the first sentence.
Author action:
Line 36, 37:Thank you for your suggestion. We delete repeated words.
Concern # 4:
Line 60: What is KF? Please provide its full name for the first-time use
Author action:
Line 62:We gave it a full name.kalman filter(KF)
Concern # 5:
Line 69: line segment tracking -> LT
Author action:
Line 71:We modified it. line tracking->LT
Concern # 6:
Figure 1: The authors need to include the definition of all symbols, i.e., theta, .., of the figure in the caption.
Author action:
Line 118-120:We supplement the definition of all symbols in Figure 1.
Concern # 7:
Line 122: cannot find alpha in Fig. 1
Author action:
We delete the definition of alpha in the figure, which is defined in the following formula.
Concern # 8:
Line 323: delete “for”
Author action:
Line 368:Thank you for your suggestion. We modified it.
Concern # 9:
Line 328: delete “there is”
Author action:
Line 373:Thank you for your suggestion. We modified it.
Concern # 10:
Line 373: Same as line 328, delete “there is”
Author action:
Line 418:Thank you for your suggestion. We modified it.
Concern # 11:
11: Line 388: fit -> fits
Author action:
Line 433:Thank you for your suggestion. We modified it.
Concern # 12:
12: Line 396: Not sure what this sentence is used for.
Author action:
Line 441:We delete this sentence, thank you for pointing out the error.
Concern # 13:
13: Section 5: The author should include all important conclusions in this section and the abstract. This section and abstract seem to emphasize different points. For example, the abstract highlights the similarity of the new method compared to the other methods, while the conclusion mostly focuses on the achieved accuracy.
Author action:
Line 489-499:We extended the conclusions and discussed them.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Authors present in the revised version of the paper the analysis of an adaptive threshold line segment feature extraction algorithm for laser radar scanning environment. Authors accepted just some of my previous comments and most of them ignored. The changes of this version in comparison with previous one is just in few details and because they also changed the references randomly (they replaced one reference with another one to make zero changes in paper), the authenticity of the paper is questionable. For that reason, I do not accept this paper in this form and I suggest revising the article into a reliable one and correcting the insufficiencies I mentioned before and are noticed below.
Page n, Row n: Authors did not respect the meaning of variables, functions, units representation. In the whole text, equations, graphs, schemes the variables, functions, lower/upper indices … are written randomly. Please, use the proper denotation because it is unreadable, what author would like to present.
Page n, Row n: Add units for variables (after expression definition).
Page n, Row n: Use minus sign in the proper places not hyphen.
Page n, Row n: Chapter titles and sentences should begin with an uppercase letter.
Page n, Row n: Authors randomly marked some parts in yellow but in general, they changed different parts of the document. What the reason was to mark different parts those were not changed?
Page n, Row n: Unify the equations size and size of expressions/equations in text, because they are copy-pasted from another document and they do not correspond to paper template.
Page n, Row n: Correct text part according to template and do not connect words, numbers and signs into one glued word.
Page 5, Row 170: What represents x and y-axes? What is legend and what is unit? Similarly other graphs.
Page 7, Row 249: What type of symbols used authors? Is it apostrophe, comma, top line, or what?
Page 9, Row 290: Description of Fig. 6 is insufficient.
Page 12, Row 391: What represents 8G? Is it the 8 GB or prepared new 8G connection type?
Page 17-18, Row n: Why authors replaced approximately half of references (one by one)?
Page 17-18, Row n: It is not possible to identify some references. Add additional information to references (authors, paper title, doi, volume, link, …) and align them according to template.
Incorporating all these comments is a necessary condition for accepting the paper.
Author Response
Concern # 1:
Page n, Row n: Authors did not respect the meaning of variables, functions, units representation. In the whole text, equations, graphs, schemes the variables, functions, lower/upper indices … are written randomly. Please, use the proper denotation because it is unreadable, what author would like to present.
Author action:
We re-edited the formula in this paper.
Page 3, Row 127:Np changed to N.
Concern # 2:
Page n, Row n: Add units for variables (after expression definition).
Author action:
Page 3, Row 119-121
Page 4, Row 144
Page 6, Row 215
We added units to the variables that first appeared in this paper.
Concern # 3:
Page n, Row n: Use minus sign in the proper places not hyphen.
Author action:
Page 4, Row 141、146:We 've re-edited the formula with the minus sign in the paper.
Concern # 4:
Page n, Row n: Chapter titles and sentences should begin with an uppercase letter.
Author action:
Page 9, Row 307:We modified the letters in the title without uppercase.
Concern # 5:
Page n, Row n: Authors randomly marked some parts in yellow but in general, they changed different parts of the document. What the reason was to mark different parts those were not changed?
Author action:We made minor adjustments to the grammar of the relevant content, marking it yellow, but there was no explanation at the time, causing misunderstanding.
Concern # 6:
Page n, Row n: Unify the equations size and size of expressions/equations in text, because they are copy-pasted from another document and they do not correspond to paper template.
Author action:
We reedited all the formulas and variables and unified the size.
Concern # 7:
Page n, Row n: Correct text part according to template and do not connect words, numbers and signs into one glued word.
Author action:
Page 10, Row 337:We made statement adjustments to i + N / 2-th similar expressions in the original text.
Concern # 8:
Page 5, Row 170: What represents x and y-axes? What is legend and what is unit? Similarly other graphs.
Author action:
Page 4, Row 161-171:The x-axis and y-axis are polar coordinates transformed into rectangular coordinates, and the unit is m. We supplement them in the article, and thank you for your suggestion.
Concern # 9:
Page 7, Row 249: What type of symbols used authors? Is it apostrophe, comma, top line, or what?
Author action:
Page 7, Row 257-258: We replace it with comma separation variable.
Concern # 10:
Page 9, Row 290: Description of Fig. 6 is insufficient.
Author action:
Page 9, Row 288-298:We replaced Figure 6, deleted the curve of the intermediate variable and gave a supplementary description.
Concern # 11:
Page 12, Row 391: What represents 8G? Is it the 8 GB or prepared new 8G connection type?
Author action:
Page 12, Row 402:We changed to 8GB.
Concern # 12:
Page 17-18, Row n: Why authors replaced approximately half of references (one by one)?
Author action:
1.Yuan H , Hao Y G , Liu J M , et al. Research on Multi-sensor Image Matching Algorithm Based on Improved Line Segments Feature[J]. Itm Web of Conferences, 2017, 11:05001.
1.Yuan Hui, Hao Ying-guang,Liu Jun-min. Research on Multi-sensor Image Matching Algorithm Based on Improved Line Segments Feature[J]. ITM Web of Conferences,2017,11:05001.
11.Ng C C , Yap M H , Costen N , et al. Wrinkle Detection Using Hessian Line Tracking[J]. IEEE Access, 2015, 3:1079-1088.
11.Choon-Ching Ng,Moi Hoon Yap,Nicholas Costen,Baihua Li. Wrinkle Detection Using Hessian Line Tracking.[J]. IEEE Access,2015,3:
Just like the above two references, when we modify the references, we replace the reference search website, in fact they are the same reference.Seven Chinese articles were included in the first edition. They were modified when they thought they could not be put in English articles. Now the revision of this edition has restored it and restored the replaced literature.
Concern # 13:
Page 17-18, Row n: It is not possible to identify some references. Add additional information to references (authors, paper title, doi, volume, link, …) and align them according to template.
Author action:
We revised the format of references.
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
Author presents in the second revised version of the paper the analysis of an adaptive threshold line segment feature extraction algorithm for laser radar scanning environment. Authors accepted mostly all of my previous comments (some of my comments authors marked as addressed, but in general, they did not realized any change or significant modification). The paper is readable, has compact structure and is written according to prescribed form. Therefore, I suggest accepting this paper in the present form.
