Prompt Mapping Tree Positions with Handheld Mobile Scanners Based on SLAM Technology
, Julián Tomaštík 2, František Chudý 2, Daniel Tunák 2, Michal Skladan 1, Ján Tuček 2 and Martin Mokroš 1,3,4,5Round 1
Reviewer 1 Report (New Reviewer)
Comments and Suggestions for AuthorsThe study compares 2 handheld laser scanners in terms of positioning accuracy of the trees. This is a revised or resubmitted version. I have not had the manuscript for review before. The question is interesting and important for the scientific community. The Stonex scanner in particular is new on the market and it is interesting to see how it performs in the forest. The problem is that the authors did not pay sufficient attention to the previous Reviews or did not incorporate suggestions. There are still many methodological ambiguities. This makes it difficult to follow the context of the paper. The statistical tests are also partly misapplied, misinterpreted or inappropriate tests.
It is particularly annoying that there are still a large number of "Error! Reference source not found" in the paper. Reviewer 4 already noted this in the previous review round. How can this happen? If I am given the opportunity for a revised version or the option to resubmit, I try even harder to deliver a perfect paper. This also applies to numerous sloppy mistakes (e.g. duplicate words, sentences, etc.) in the manuscript.
Nonetheless, I believe that the paper has potential and the underlying data is good. The authors must make an effort to find the appropriate methods and, above all, to describe them appropriately. I would consider an additional advantage of the paper if the data (reference data, point clouds, results of automatic tree finding and measurement, etc) were published.
I suggest intensive major revisions.
Detailed comments can be found in the attached pdf.
Comments for author File: 
Comments.pdf
Author Response
The study compares 2 handheld laser scanners in terms of positioning accuracy of the trees. This is a revised or resubmitted version. I have not had the manuscript for review before. The question is interesting and important for the scientific community. The Stonex scanner in particular is new on the market and it is interesting to see how it performs in the forest. The problem is that the authors did not pay sufficient attention to the previous Reviews or did not incorporate suggestions. There are still many methodological ambiguities. This makes it difficult to follow the context of the paper. The statistical tests are also partly misapplied, misinterpreted or inappropriate tests.
Answer of authors: Thank you for your recommendations; your objective comments have helped us elevate the contribution to a higher level. Edits in the text have been made in the change mode to ensure they are clearly visible.
It is particularly annoying that there are still a large number of "Error! Reference source not found" in the paper. Reviewer 4 already noted this in the previous review round. How can this happen? If I am given the opportunity for a revised version or the option to resubmit, I try even harder to deliver a perfect paper. This also applies to numerous sloppy mistakes (e.g. duplicate words, sentences, etc.) in the manuscript.
Answer of authors: Thank you for pointing out and highlighting the shortcomings in the text. To ensure a flawless submission next time, we'll thoroughly review the references and content, searching for duplications or errors. We aim to provide a revised version that meets the highest standards and meticulously addresses all these concerns. We intend to send the contribution to a language proofreader before potential publication, aiming to resolve any language or grammatical deficiencies.
Nonetheless, I believe that the paper has potential and the underlying data is good. The authors must make an effort to find the appropriate methods and, above all, to describe them appropriately. I would consider an additional advantage of the paper if the data (reference data, point clouds, results of automatic tree finding and measurement, etc) were published.
In the Data Availability Statement section, we state: The data presented in this study are available on request from the corresponding author. The data are not publicly available due to large volume of data.
Comments from pdf:
Page1
Line 2 ** insert "of"
Answer of authors: Thank you for the recommendation, we will wait for proofreading
Line 20** This is too imprecise for the abstract. Name the factors.
Answer of authors: We agree and added the factors influencing the accuracy of positioning in HMLS Simultaneous Localization and Mapping-aided solutions were defined, considering the scanner type, distance from the trajectory, forest structure, tree species, and Diameter at Breast Height.
Page 2
line 84** You can't do it like that. The quote should be inserted as follows: what described in Chen et al [20]. Please check everywhere and correct if necessary.
Answer of authors: Supplemented according to the jorunal requirements
Line 86** SLAM comes from robotics - right. But here it reads as if SLAM is a process in robotics. Yes, but not only in robotics - also in transportation, industry, or even in HMLS. This needs to be explained in more detail in 1-2 more sentences. So you would think that SLAM only exists in connection with robots - that's not true. –
Answer of authors: Thank you added
Page 3
Pine 120** What is TORCH - what does it mean in full? That needs to be explained.
Answer of authors: This is not a essential part of the manuscipt, if the reader is interested, the information can be found in the article we are quoting.
Line 131** That's right! And that's why the advantages of the study need to be better highlighted. I would also pay more attention to the comparison of the two scanners. I would also publish the associated data sets (point clouds, reference data, etc.) from the study. This would be very important for the community and would also make the study more valuable.
Answer of authors: Advantages added; In the Data Availability Statement section, we state: The data presented in this study are available on request from the corresponding author. The data are not publicly available due to large volume of data.
Line 147** position errors?
Answer of authors: Thanks added
page 4
line 154** why only flat terrain?
Answer of authors: We chose ideal conditions for data collection; the research obviously needs to be expanded in the future. We include a section on the potential effects of this factor on the data in the discussion. The important factor is not the slope (terrain). Important is tree species and and the forest density, these factors have been included.
Line 158** year?
Answer of authors: Thanks added
Line 161** Where exactly is the tree position (i.e. x/y)? There are eccentric/unshaped cross-sections where you cannot simply take dbh/2. This must at least be discussed. One could also have taken TLS data and TLS cross-sections and their center of gravity e.g. as center and reference position.
Answer of authors: we consider the total station to be the best option for collecting data on positions, since we measure with millimeter accuracy.
Line178** How did you come up with the 3 cm? How was this evaluated? Why were the trees only mapped with regard to x/y? The z coordinate would also have been interesting for the evaluation.
Answer of authors: Added 3cm threshold information; We did not use the Z coordinate because we do not see the point in its determination. In order to calculate all indices, it is necessary to know the X and Y coordinates
Line 179** Something like "Error! Reference source not found..." must not happen. Especially not because it has already been noted by a reviewer in one of the previous rounds of reviewers. This makes the paper illegible and you can no longer follow the methods, explanations etc. This must not happen in a resubmission! –
Answer of authors: corrected.
Page 5
Line 189** I would leave out the scientific names of the tree species and just write the English ones. That way the lines will be a little less high.
Answer of authors: There has not been such a request from any reviewer so far, so we will ignore it at this stage.
Page 6
Line 200** Actually, everything comes from the manufacturer's specification. Or have you tested the range yourself, or even the protection class yourself? ;-) etc. –
Answer of authors : Right :D thanks for pointing out.
Line203** How does that look like? Where does each scanner make exactly 0/0 or how does the scanner know exactly where the GCP is? This is very, very crucial for the study, because all accuracy measures then depend on it.
maby you can make a figure.
Answer of authors: In the case of both devices, movement is required for the proper functioning of the SLAM algorithm. In the case of the GeoSLAM device, if the desired stop of the operator on the GCP occurs for approximately 10 seconds, this is a signal for the processing algorithm to assign concrete coordinates for the location and concrete time period of the measurement. In the case of the Stonex device, the signal about the measurement of the GCP point comes directly from the operator who is in the field and starts the manual measurement of the GCP point in the application.The correct location of the coordinates is ensured by a developer-defined specific shift in the direction of the XYZ axis, which is taken into account by the calculation algorithm during the post-processing of the measurement, which ensures the exact location of the GCP point and the transformation of the point cloud.
Line 213** with the help of the plate? That is important! If you are a few centimeters off, it has a big impact. -- Answer of authors: Answer in previous comment.
Line 224* Georeferencing is usually a very difficult step in HMLS and usually very time-consuming. Please provide more details.
Answer of authors: We proceeded on the basis of : https://knowledge.faro.com/Software/GeoSlam/GeoSLAM_Draw/Georeferencing_Process_with_GeoSLAM_Draw
Line 227** The GeoSLAM can also be equipped with an RGB camera. Why is only the Stonex camera described? RGB has nothing to do with the study anyway. Either describe both or none at all.
Answer of authors: At stonex, we used an RGB camera, not at GeoSlam.
Line 234** The workflow for GeoSLAM and Stonex must be shown at least in tabular form or in a graphic, at least schematically. A scientific study requires that the data/results are reproducible and that the methods can be applied! And as it is now, it cannot be reproduced and the method cannot be applied.
Answer of authors: We requested the procedure directly from the manufacturer
https://knowledge.faro.com/Software/GeoSlam/GeoSLAM_Draw/Georeferencing_Process_with_GeoSLAM_Draw
Page 8
Line 245** Thousands separator
How do you get 5141 trees? That has nothing to do with the values in Table 1. Please clarify!
Answer of authors: Added and explained.
Line 253** mean, median of what? mean error - would be the bias. Describe better!
Answer of authors: citation from: https://www.mdpi.com/2072-4292/13/22/4677
Line 247** 10 cm RMSE with the dbh estimate is a lot. I know it's not about dbh estimation in the paper. But, if the DBH is bad, the cross section and thus the x/y of the single tree should also be bad.
Or does the data come from another study? This needs to be described better/clearer - it's impossible to follow! -- Answer of authors: Added; citation reformatted.
Line 253** Why is there now an "old" study? What are they trying to say? Something like that has no place in the methods. You can make comparisons with other studies in the introduction or in the discussion.
Answer of authors: Rewrited.
Line 258** The same here. That doesn't belong in the methods. It only confuses and makes it unreadable. What does the TLS/MLS comparison have to do with the specific study?
Answer of authors: Rewrited.
Line 264** Why does a focus shift? A study should have clear objectives. These should be defined at the end of the introduction. The methods should explain how to achieve the objectives. And then show in the results what has come out and then critically discuss and compare this in the discussion.
It should not (as here) suddenly be stated in the methods that the focus has shifted - nobody knows anything about it anymore!
Answer of authors: Wrong word choice – changed.
Page 9
Line 278** You need an example, or an equation or a figure. Otherwise I don't understand what you have done. What if the trajectory also has an error? You are evaluating the distances tree to trajectory - right? Now, according to my experience, it also happens that trajectories can also have a drift. How do you deal with this in the study? -- Answer of authors: We compared the actual trajectory from the schematics in Qgis.
Line 278** ?? position error computed as diff between diameters?
Answer of authors: Mistake, thank you for pointing out; Corrected.
Line 282** also yref and yest - right?
Answer of authors: Of course; added.
Line 291** Better explanation needed. Presentation of test statistics necessary. I wonder how you can do a Kruskal test or a hypothesis test at all. For example, if you want to test whether the position errors are different in terms of N/ha, then you have 5 groups with one observation each. I can't get a hypothesis test together with the data or the test outputs are then meaningless. Please clarify.
Answer of authors: since it is a categorical variable (tree species), we used a common statistical procedure
Line 296** From my point of view, it should be a paired t-test, because GeoSLAM and Stonex were evaluated on the same plot, with the same trees.
Answer of authors: in this case this test has better strength
Line 298** The half-sentence contains reduntante information. The sentence above it already says the same thing. Answer of authors: Corrected.
Line 300** I don't understand the sentence. Do you want to see if there is a correlation between the distance (tree trajectory) and the error of the single tree position? If so, then this must be written more clearly - nobody understands it as it is now.
Answer of authors: Rewrited.
Line 303** From now on I won't look at everything closely anymore. There are too many questions and ambiguities in the methods. They must first be cleaned up, then adjusted accordingly to the results and discussion.
Below you will only find a few comments and they do not claim to be complete.
Answer of authors: Thank you for your recommendations.
Page 10
Line 312** ?
Answer of authors: Rewrited.
Line 323** I can't see that in Table 3.
Answer of authors: Corrected.
Page 11
Line 347** This has to be described in the methods with examples, equations, illustrations etc. That's not how you understand it.
Answer of authors: Corrected
Line 355** This must be explained and described in the methods and not in the results!
Answer of authors: Rewrited.
Line 376** Units are missing
Answer of authors: Added.
Page 12
Line 384** As already written in the methods. It doesn't work that way and it's also not statistically correct.
Answer of authors: the only possible statistical method for a categorical variable where there are more than two categories
Line 391** I do not understand that. Describe more clearly and, above all, describe better in the methods.
Answer of authors :I don't think it is necessary to describe statistical methods
Page 13
Line 436** This is a discussion and not a result!
Answer of authors : paragraph replaced.
Line 488** Pearson’s correlation coefficient can never say anything about significance because it is not a hypothesis test. All you get is a correlation. It can be high or low - but that's it.
Answer of authors: Ok thanks deleted.
Line 467** That's double. The same thing is already in lines 454-457. How can this happen? The paper will be resubmitted and should have undergone several review/editing stages by the authors. Such sloppy errors can be found scattered throughout the manuscript.
Answer of authors :Duplicate removed.
Line 478** year?
Answer of authors: Added.
Line 497** GeoSLAM already has the ZEB Horizon RT (Real Time). Where you can start, finish and view on the fly via the app.
Answer of authors: Yes, we know, but we worked with an older model, which is available from the GeoSLAM distributor in Slovakia.
Page 17
Line 565** ?
Answer of authors :Corrected.
Page 18
Line 573 -- well written!
Answer of authors Thank you!
Line 593** ?
Answer of authors : [citation 64] Description of the positioninng in tropial forest with dense canopy.
Line 596** RMSE
Answer of authors :Corrected.
Line 608** What does the sentence mean?
Answer of authors : Rewrited.
Line 643** ?
Answer of authors :Corrected
Line 685** double sentence parts!
Answer of authors :Corrected.
Reviewer 2 Report (New Reviewer)
Comments and Suggestions for Authors
---
Subject: Review Report for Manuscript "Prompt Mapping Tree Positions with Handheld Mobile Scanners Based on SLAM Technology" Submitted to Land Journal
Dear Authors,
I have reviewed the manuscript titled "Prompt Mapping Tree Positions with Handheld Mobile Scanners Based on SLAM Technology" submitted to Land Journal. Below are my observations and suggestions for improvement:
- **Abstract**: The abstract is well-structured, yet I recommend supplementing it with more pertinent information about the manuscript's content.
- **Introduction Section**: I propose replacing hyperlinks with formal web references in the introduction for better scholarly presentation.
- **Citation Errors**: There are numerous citation errors throughout the manuscript that necessitate careful review and correction.
- **Figure 2**: It would significantly enhance clarity and precision to include a scale bar for Diameter at Breast Height (DBH) values in Figure 2.
- **Line 218**: Clarification is needed regarding the significance of the measurement scale mentioned—6 mm @ 100 m.
- **Line 343**: Please clarify if the reference pertains to Figure 1 or Figure 5. Additionally, consider splitting Figure 1 for clearer representation of values beyond 26.
- **Table 4**: Kindly add units to Table 4 for better comprehension.
- **Discrepancy between Figures and Table**: I found a discrepancy between Figure 1 and Table 4. The figure suggests no trees beyond 30 meters, but table 4 depicts a tree near 60 meters from the trajectory.
- **Table 5**: The findings in Table 5 are particularly intriguing. I anticipated higher DBH errors with distant trees due to lidar scattering and noise effects.
-**Table numbers**: the table numbers should be checked.
- **Discussion Section**: It might be beneficial to relocate and condense several parts of the discussion into the methods and results sections to streamline the content.
In summary, while the paper holds promise and interest, revisions are necessary to ensure accuracy and clarity across various sections.
Bests.
---
Author Response
- **Abstract**: The abstract is well-structured, yet I recommend supplementing it with more pertinent information about the manuscript's content.
Answer of authors : Thank you. As you know the abstract is limited by the number of words, so we tried to formulate it so that it contains the best possible summary of the described issue in accordance with general standards.
- **Introduction Section**: I propose replacing hyperlinks with formal web references in the introduction for better scholarly presentation.
Answer of authors : Thanks
- **Citation Errors**: There are numerous citation errors throughout the manuscript that necessitate careful review and correction.
Answer of authors : Thank you for pointing out and highlighting the shortcomings in the text. To ensure a flawless submission next time, we'll thoroughly review the references and content, searching for duplications or errors. We aim to provide a revised version that meets the highest standards and meticulously addresses all these concerns. We intend to send the contribution to a language proofreader before potential publication, aiming to resolve any language or grammatical deficiencies.
- **Figure 2**: It would significantly enhance clarity and precision to include a scale bar for Diameter at Breast Height (DBH) values in Figure 2.
Answer of authors: it is only a scheme and all data are in the form of numbers
- **Line 218**: Clarification is needed regarding the significance of the measurement scale mentioned—6 mm @ 100 m.
Answer of authors: modified
- **Line 343**: Please clarify if the reference pertains to Figure 1 or Figure 5. -- OK Additionally, consider splitting Figure 1 for clearer representation of values beyond 26.
Answer of authors : Thank you for the recommendation, we would like to leave the figure as it is.
- **Table 4**: Kindly add units to Table 4 for better comprehension.
Answer of authors: modified
- **Discrepancy between Figures and Table**: I found a discrepancy between Figure 1 and Table 4. The figure suggests no trees beyond 30 meters, but table 4 depicts a tree near 60 meters from the trajectory.
Answer of authors: We hope we understood your comment as you meant it. Of course, the trees were also located beyond the border of our research areas since they are wide forest stands. However, we visualized the data that we collected manually in the field - trees and their attributes inside the research area. In order to avoid misunderstanding, we modified the name of Figure 2 (Figure 2. Research plots – situation maps based on reference data measurements).
- **Table 5**: The findings in Table 5 are particularly intriguing. I anticipated higher DBH errors with distant trees due to lidar scattering and noise effects.
Answer of authors : Yes, it is clear from our and from previous research in our author's collective that hmls technologies achieve very good results in determining DBH. For your interest: colleagues published interesting research The Comparison of Stem Curve Accuracy Determined from Point Clouds Acquired by Different Terrestrial Remote Sensing Methods - where the input was tree thicknesses at different heights above the terrain and their high accuracy was also demonstrated.
-**Table numbers**: the table numbers should be checked.
Answer of authors: modified
- **Discussion Section**: It might be beneficial to relocate and condense several parts of the discussion into the methods and results sections to streamline the content.
Answer of authors :Modifications and corrections have been made and highlighted.
In summary, while the paper holds promise and interest, revisions are necessary to ensure accuracy and clarity across various sections.
Bests.
Dear Reviewer,
Thank you for your recommendations; your objective comments have helped us elevate the contribution to a higher level. Edits in the text have been made in the change mode to ensure they are clearly visible.
Have a nice day.
Round 2
Reviewer 2 Report (New Reviewer)
Comments and Suggestions for AuthorsAcceptable
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors1. introduction: there are many handheld mapping system, i.e., finnish geospatial research institute has many papers on laser scanning;
Author Response
The comments were accepted, sources were added to the Introduction section.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsIn this study, the authors used two different handheld laser scanning devices, the GeoSLAM ZEB Horizon and the Stonex X120GO SLAM HMLS, to compare the positional accuracy of their scans in different forest environments, and to evaluate their positioning accuracy by comparing the error of the estimated diameter with the reference diameter. The study also analyzed the association of factors such as tree and device trajectory, forest structure, tree species, trunk DBH, etc. with scanning positional errors to identify factors that may affect the error.
1. In the abstract, especially the lack of a value description of the results at the end of the abstract, it is recommended to add a few words about the affirmation and sublimation of the value of the research results.
2. The authors compared the performance and mean position RMSE of the GeoSLAM ZEB Horizon and the Stonex X120GO SLAM. However, it is unclear whether the influence of different terrain factors on the data collected by HMLS was taken into account when selecting the experimental data plot. It is recommended that the authors clearly state in the article whether the choice of experimental data plot took into account the difference in terrain steepness and analyze it. If the effect of terrain steepness is not considered, the authors can mention this and discuss the potential impact of this choice on the results of the study.
3. It is mentioned in the article that the collection time is from March 2022 to November 2022, but it is not clear whether the impact of leaf-off and leaf-on in different seasons in European forests on data collection is considered. It is recommended that the authors explicitly discuss in the article whether the impact of different seasons is considered and explain why this time period was chosen for data collection. If the impact of seasons is considered, the reasons for choosing this time period can be described in detail and explain why this time period can represent the conditions of leaf-off and leaf-on. If the impact of seasons is not considered, the authors can mention this and discuss the potential impact of this choice on the results of the study.
4. In Figure 2. Research plots - situation maps. and Table 1. Research plots summary information. The difficulty level of each collected plot should be described so that the reader can more clearly understand the effect of HMLS in different application scenarios. In Figure 2, some additional information can be added next to the annotation of each plot, such as number of trees, tree density, etc. In this way, the reader can visually understand the tree condition of each plot and compare it with the performance of HMLS. Different colors or symbols can be used to represent the number and density of trees in different plots. In Table 1, a column can be added or an existing column can be modified to describe the difficulty level of each plot. For example, information such as number of trees, tree density, etc. can be added to more fully describe the characteristics of each plot. This information can help the reader understand the tree condition of each plot and have a clearer understanding of how HMLS performs on different plots.
5. When the author explores the influence of DBH on the position error, the average DBH is more than 20 cm, and the measured position point of GNSS is the point around the trunk, and the tree position usually obtained is often the center of the tree. The size will affect the position of the measured GNSS. By comparing the corresponding tree position RMSE, this part should also need to be considered.
6. P82-83, P455 lack literature support.
7. The introduction mentions the importance of tree position research, and also mentions some measurement methods about traditional position determination methods and SLAM technology. However, there is a lack of forestry positioning descriptions for other technologies (such as acousto-optic technology, etc.). It is recommended that the author increase Relevant literature review.
8. Errors are indicated in the paper at P120, P135, P142, P150, P162, P170, P256, P269, P303, P389, P392, P400, P428, P432, P444, P466, P475, etc.
9. Section 9.2.1, on the evaluation of accuracy, does not fully explain the factors involved, and there is a lack of indicators on R2. ...
10. In general, the author compares the comparison of the tree position accuracy of the two scanners, but concludes that no clear conclusions are drawn and the shortcomings of this study are suggested. It is recommended to further add improvements about this study in the discussion section.
11. The conclusion part is too cumbersome, and it is recommended that the author further condense and express the research conclusions clearly.
Author Response
In this study, the authors used two different handheld laser scanning devices, the GeoSLAM ZEB Horizon and the Stonex X120GO SLAM HMLS, to compare the positional accuracy of their scans in different forest environments, and to evaluate their positioning accuracy by comparing the error of the estimated diameter with the reference diameter. The study also analyzed the association of factors such as tree and device trajectory, forest structure, tree species, trunk DBH, etc. with scanning positional errors to identify factors that may affect the error.
In the abstract, especially the lack of a value description of the results at the end of the abstract, it is recommended to add a few words about the affirmation and sublimation of the value of the research results.
Following recommendations, the Abstract was reformulated.
The authors compared the performance and mean position RMSE of the GeoSLAM ZEB Horizon and the Stonex X120GO SLAM. However, it is unclear whether the influence of different terrain factors on the data collected by HMLS was taken into account when selecting the experimental data plot. It is recommended that the authors clearly state in the article whether the choice of experimental data plot took into account the difference in terrain steepness and analyze it. If the effect of terrain steepness is not considered, the authors can mention this and discuss the potential impact of this choice on the results of the study.
Comments are included in the Material and Methods, the description of the study area and the Discussion. The areas were chosen so that they were flat, we chose ideal conditions for data collection, the research obviously needs to be expanded in the future. We included a section on the potential effects of this factor on the data in the discussion.
. It is mentioned in the article that the collection time is from March 2022 to November 2022, but it is not clear whether the impact of leaf-off and leaf-on in different seasons in European forests on data collection is considered. It is recommended that the authors explicitly discuss in the article whether the impact of different seasons is considered and explain why this time period was chosen for data collection. If the impact of seasons is considered, the reasons for choosing this time period can be described in detail and explain why this time period can represent the conditions of leaf-off and leaf-on. If the impact of seasons is not considered, the authors can mention this and discuss the potential impact of this choice on the results of the study.
Comments are included in the Material and Methods and the Discussion.
. In Figure 2. Research plots - situation maps. and Table 1. Research plots summary information. The difficulty level of each collected plot should be described so that the reader can more clearly understand the effect of HMLS in different application scenarios. In Figure 2, some additional information can be added next to the annotation of each plot, such as number of trees, tree density, etc. In this way, the reader can visually understand the tree condition of each plot and compare it with the performance of HMLS. Different colors or symbols can be used to represent the number and density of trees in different plots. In Table , a column can be added or an existing column can be modified to describe the difficulty level of each plot. For example, information such as number of trees, tree density, etc. can be added to more fully describe the characteristics of each plot. This information can help the reader understand the tree condition of each plot and have a clearer understanding of how HMLS performs on different plots.
Following recommendations, Figure 2 and Table 1 have been updated to include additional information.
Thank you for the warning; we realized that we had not adequately described the collection of reference data. We have now added an explanation in the Material and Methods section.
- P82-83, P455 lack literature .
Additional literary sources have been added.
- introduction mentions the importance of tree position research, and also mentions some measurement methods about traditional position determination methods and SLAM technology. However, there is a lack of forestry positioning descriptions for other technologies (such as acousto-optic technology, etc.). It is recommended that the author increase Relevant literature review.
In our work, we only deal with ground technologies of close range, which enable us to measure DBH with high portability.
- Errors are indicated in the paper at P120, P135, P142, P150, P162, P170, P256, P269, P303, P389, P392, P400, P428, P432, P444, P466, P475, .
Thanks for the notice, the errors have been removed.
- , does not fully explain the factors involved, and there is a lack of indicators on R2.
We believe the factors have been described sufficiently, considering your previous recommendations that we have incorporated into the text.
. In general, the author compares the comparison of the tree position accuracy of the two scanners but concludes that no clear conclusions are drawn and the shortcomings of this study are suggested. It is recommended to further add improvements about this study in the discussion section.
Comments are included in the Introduction and the Discussion section.
. The conclusion part is too cumbersome, and it is recommended that the author further condense and express the research conclusions clearly.
The Conclusion section was reformulated.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript titled "Prompt Mapping Tree Positions with Handheld Mobile Scanners Based on SLAM Technology" aims to assess the performance of two handheld mobile laser scanners in mapping tree positions. The authors conducted tests on the accuracy of these scanners in five different plots in Slovakia. While the overall structure of the paper adheres to scientific standards, there are several issues within the content that hinder comprehension.
- Firstly, the presence of numerous "Error! Reference source not found" instances throughout the manuscript makes it challenging to follow. This creates a barrier to understanding the material.
- Secondly, inconsistencies in the presentation of methodology and results contribute to confusion and make reading time-consuming.
o For instance, the discrepancy between the mention of "588 reference trees, 535 or 90.98% were successfully detected" (L203) and later stating that "The procedure allowed the evaluation of 421 trees from 436 individual trees" raises questions about the origin of these numbers. The total number of trees calculated from Table 1 for all five plots is 457, adding to the confusion.
o Moreover, the absence of tables in sections 3.3 and 3.4, where tables were expected based on the preceding sections, further contributes to inconsistency.
- The introduction section requires revision as it falls short in justifying the research problem. While the statement "our best knowledge no one compared these two devices…" is presented, it might not suffice for a scientific inquiry that necessitates a more thorough exploration of the research problem.
- Language quality is a notable issue throughout the manuscript. Word using in some sentences may not be suitable for a scientific paper, and a comprehensive English revision is needed.
- Additionally, terms such as "DBH" are introduced without explanations or references. Methodologies like the Kruskal-Wallis test and Pearson’s correlation are mentioned without proper references, leaving readers without guidance.
Given these shortcomings, I have to reject the manuscript, judging by the current state of the manuscript.
Comments on the Quality of English LanguageLanguage quality is a notable issue throughout the manuscript. Word using in some sentences may not be suitable for a scientific paper, and a comprehensive English revision is needed.
Author Response
Firstly, the presence of numerous "Error! Reference source not found" instances throughout the manuscript makes it challenging to follow. This creates a barrier to understanding the material.
Thanks for the notice, the errors have been removed.
- Secondly, inconsistencies in the presentation of methodology and results contribute to confusion and make reading time-consuming.
Thank you for your opinion, the problematic parts have been checked and we are going to send the article for language proofreading.
o For instance, the discrepancy between the mention of "588 reference trees, 535 or 90.98% were successfully detected" (L203) and later stating that "The procedure allowed the evaluation of 421 trees from 436 individual trees" raises questions about the origin of these numbers. The total number of trees calculated from Table 1 for all five plots is 457, adding to the confusion.
In case of this study, the procedure allowed the evaluation of 417 trees from 457 individual trees. The mentioned discrepancy arose from a misinterpretation of the sentence in the text, the problematic part has been checked and rewrite.
o Moreover, the absence of tables in sections 3.3 and 3.4, where tables were expected based on the preceding sections, further contributes to inconsistency .
The statistical results are presented in the text, and we believe that separate tables are not necessary. Our claims are supported by a statistical analysis, with the main parameters such as regression coefficients or p-values provided in the text.
- The introduction section requires revision as it falls short in justifying the research problem. While the statement "our best knowledge no one compared these two devices…" is presented, it might not suffice for a scientific inquiry that necessitates a more thorough exploration of the research problem.
Following recommendations, additional literary sources have been added to the Introduction section.
- Language quality is a notable issue throughout the manuscript. Word using in some sentences may not be suitable for a scientific paper, and a comprehensive English revision is needed.
We intend to arrange for a language proofreading in the future.
- Additionally, terms such as "DBH" are introduced without explanations or references.
Thanks for the notice, the errors have been removed.
Methodologies like the Kruskal-Wallis test and Pearson’s correlation are mentioned without proper references, leaving readers without guidance.
We have included literary sources describing the mentioned tools in the manuscript.
Given these shortcomings, I have to reject the manuscript, judging by the current state of the manuscript.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsFirst of all, I would like to thank the Land Editorial Board for the opportunity to participate in this evaluation process.
The paper addresses the quality of information captured with handheld mobile scanners in forested areas in order to establish tree inventories, including tree position, as well as some basic parameters such as DBH itself. Actually, the article focuses on the comparison of two specific commercial systems, which may raise some doubts as to the specific interest of the research. However, it could be considered that the interest may be based on the procedure of how to compare the results of the two different systems, not so much on the concrete results of these systems that logically are only an example of the current high number of these systems, and that logically will be changing over time, so the concrete conclusions -beyond the procedures- will really have a very short real life, logically limiting the real impact of this publication.
In any case, it is important to consider the role that SLAM-based systems can have in order to obtain information in this type of area, which presents an undoubted complexity -visibility, potential GNSS signal losses, etc.-. However, it is important to bear in mind that if the interest is focused not so much on the systems but on the methodology, it must be especially careful, in order to guarantee its applicability in different cases. In this sense, I consider that the design of the experiment could be improved by applying some changes:
1) Perform repeatability tests of results by surveying the same plot on different occasions.
2) Analyze the influence that the type of route taken in the plot can have on the surveys. For example, in this type of systems, it is advisable to consider a circular path that allows to "close" the beginning and the end of it, in order to readjust all the errors of the path.
3) Finally, I consider that it would be advisable to carry out an analysis of the topography of the surveyed plots, it is evident that the difficulty of obtaining information is not the same in plots with steep slopes or complex orography as in flat plots.
4) The size of the plots is considered quite limited for forestry operations. What would be the error forecast in the case of working not with experimental plots, but with real plots with irregular shapes and a larger size? Would the error be constant?
5) Regarding the influences analysis section, it requires an important analysis. Problems of localization and determination of DBH are being raised. It is evident that the structure of the forest surely has an influence due to the lack of visibility, or the species -but not due to the quality of the measurement, but due to the uncertainty when establishing the "center" of the tree or the DBH-, or the DBH itself, since the smaller the diameter, the more accurate the determination of the center both by total station and by SLAM systems -in this case, the analysis is not entirely clear-.
6) In this regard, it is considered important to review the section corresponding to the discussion of the results achieved.
Although a detailed error analysis has been performed, I believe that a basic component to be considered is missing - either it has been established in a partial way, through correlations between errors and x and y coordinates, or between different systems. For this, it would be necessary to consider what is the spatial distribution of the errors, and in order to be able to establish if the errors accumulate along the path or not, or if it is related to any other parameter.
Being aware of the short time between revisions, I consider that these modifications would be of great interest, although they should surely lead to a new publication and not to a revision of this one.
Regarding formal issues:
(a) Although this is a frequently used abbreviation, taking into account the characteristics of the journal Land, it would be advisable to indicate the meaning of DBH, the first time it is cited in the manuscript (L.41).
b) The study proposes an admissible positional accuracy level of 0.5m, which should be justified. It is a question of absolute positional accuracy (with respect to ground coordinates) or relative (with respect to other elements of the same survey). Thus, the usefulness of considering absolute positional accuracy as an element of registration or inventory of trees, versus relative, for the consideration of growth patterns, densities, etc., should be justified.
c) It is necessary to comment on the specific problems of SLAM (or VisualSLAM) in this type of scenario. An inexperienced user may be surprised by the difference between the accuracy stated in the table of technical specifications of the systems and those actually achieved. It is also important to consider the influence of the potential loss of GNSS signal in this type of system.
d) Clarify how the position of each of the trees has been measured by means of the total station and GNSS (have different measurements been taken? how has it been possible to estimate the center of the tree which is understood to be the reference point in the comparison, although it is considered that the trees have a small diameter -with DBH less than 30cm-.
e) L202 ... It is not clear which are the errors reached. How is it possible that an RMSE with a negative value of -0.007m appears?
f) Taking into account the importance of the FSCT tool used, it would be advisable to comment on the methodology used to make comparisons.
g) A more in-depth analysis of the errors obtained in Table 3 (L267) is needed. In plots with conditions
On the other hand, it is necessary to indicate that it has not been possible to verify the correct citation of the bibliographic references as they appear as errors in the text (Error! Reference source not found).
Errata:
L138 ... Missing ) ... measurements).
L144 ... Plot (should be on the same line).
L154 (table 2). Relative accuracy appears in the GeoSLAM ZEB Horizon system with *** but does not appear in the table footer, only * When processing data in GeoSLAM Connect V2.
Author Response
The paper addresses the quality of information captured with handheld mobile scanners in forested areas in order to establish tree inventories, including tree position, as well as some basic parameters such as DBH itself. Actually, the article focuses on the comparison of two specific commercial systems, which may raise some doubts as to the specific interest of the research. However, it could be considered that the interest may be based on the procedure of how to compare the results of the two different systems, not so much on the concrete results of these systems that logically are only an example of the current high number of these systems, and that logically will be changing over time, so the concrete conclusions -beyond the procedures- will really have a very short real life, logically limiting the real impact of this publication.
Thank you for the intriguing idea.
However, we believe that it's crucial to work efficiently with the resources at hand. In our current time and space, obtaining a high-quality, quick, and easy image of the environment is essential. We consider it logical to provide information about available devices for specific tasks to the general public. Particularly in times of intense climate change and increasing disturbances, it's vital to rapidly capture high-quality environmental data. This data can support the protection of special areas and ecosystems and be used for modelling different scenarios. This enables us to respond promptly and effectively to the changes society faces, which are often unavoidable.
In any case, it is important to consider the role that SLAM-based systems can have in order to obtain information in this type of area, which presents an undoubted complexity -visibility, potential GNSS signal losses, etc.-. However, it is important to bear in mind that if the interest is focused not so much on the systems but on the methodology, it must be especially careful, in order to guarantee its applicability in different cases.
Comments are included in the Introduction and Result sections.
In its essence, SLAM technology does not need a GNSS signal for its operation, that is the reason why it is advantageous to apply it even in an environment where GNSS is absent. On the other hand, in our work we presented a methodology that is a kind of geodetic basis (transformation into a coordinate system using 4 known points) in combination with SLAM technology in a complex environment - thus we demonstrate that it is possible to achieve interesting results with good positional accuracy in an affordable way.
Formulation and Structure of the SLAM Problem
SLAM is a process by which a mobile robot can build a map of an environment and at the same time use this map to deduce its location. In SLAM, both the trajectory of the platform and the location of all landmarks are estimated online without the need for any a priori knowledge.
In this sense, I consider that the design of the experiment could be improved by applying some changes:
1) Perform repeatability tests of results by surveying the same plot on different occasions.
2) Analyze the influence that the type of route taken in the plot can have on the surveys. For example, in this type of systems, it is advisable to consider a circular path that allows to "close" the beginning and the end of it, in order to readjust all the errors of the path.
3) Finally, I consider that it would be advisable to carry out an analysis of the topography of the surveyed plots, it is evident that the difficulty of obtaining information is not the same in plots with steep slopes or complex orography as in flat plots.
4) The size of the plots is considered quite limited for forestry operations. What would be the error forecast in the case of working not with experimental plots, but with real plots with irregular shapes and a larger size? Would the error be constant?
5) Regarding the influences analysis section, it requires an important analysis. Problems of localization and determination of DBH are being raised. It is evident that the structure of the forest surely has an influence due to the lack of visibility, or the species -but not due to the quality of the measurement, but due to the uncertainty when establishing the "center" of the tree or the DBH-, or the DBH itself, since the smaller the diameter, the more accurate the determination of the center both by total station and by SLAM systems -in this case, the analysis is not entirely clear-.
6) In this regard, it is considered important to review the section corresponding to the discussion of the results achieved.
Although a detailed error analysis has been performed, I believe that a basic component to be considered is missing - either it has been established in a partial way, through correlations between errors and x and y coordinates, or between different systems. For this, it would be necessary to consider what is the spatial distribution of the errors, and in order to be able to establish if the errors accumulate along the path or not, or if it is related to any other parameter.
Being aware of the short time between revisions, I consider that these modifications would be of great interest, although they should surely lead to a new publication and not to a revision of this one.
Thank you for your comments, we will definitely consider them when creating the next experiment.
Regarding formal issues:
(a) Although this is a frequently used abbreviation, taking into account the characteristics of the journal Land, it would be advisable to indicate the meaning of DBH, the first time it is cited in the manuscript (L.41).
Thanks for the notice, the errors have been removed.
The study proposes an admissible positional accuracy level of 0.5m, which should be justified. It is a question of absolute positional accuracy (with respect to ground coordinates) or relative (with respect to other elements of the same survey). Thus, the usefulness of considering absolute positional accuracy as an element of registration or inventory of trees, versus relative, for the consideration of growth patterns, densities, etc., should be justified.
The entire study deals with the question of absolute positional accuracy. As mentioned above, in our work we presented a methodology transformation into a coordinate system using 4 known points - from local to a specific coordinate system. Comments are included in the Material and methods section.
- c) It is necessary to comment on the specific problems of SLAM (or VisualSLAM) in this type of scenario. An inexperienced user may be surprised by the difference between the accuracy stated in the table of technical specifications of the systems and those actually achieved. It is also important to consider the influence of the potential loss of GNSS signal in this type of system.
Comments are included in the Introduction section.
- d) Clarify how the position of each of the trees has been measured by means of the total station and GNSS (have different measurements been taken? how has it been possible to estimate the center of the tree which is understood to be the reference point in the comparison, although it is considered that the trees have a small diameter -with DBH less than 30cm-.
Comments are included in the Material and Methods section.
- e) L202 ... It is not clear which are the errors reached. How is it possible that an RMSE with a negative value of -0.007m ?
Cited from: Krisanski, S.; Taskhiri, M.S.; Gonzalez Aracil, S.; Herries, D.; Muneri, A.; Gurung, M.B.; Montgomery, J.; Turner, P. Forest Structural Complexity Tool—An Open Source, Fully-Automated Tool for Measuring Forest Point Clouds. Remote Sens. 2021, 13, 4677. https://doi.org/10.3390/rs13224677
“A comparison of the manual reference DBH against the automatically extracted measurements is visualised in Figure 11. Mean, median and RMS DBH errors were −0.007 m, 0.008 m and 0.072 m respectively. Of the 588 reference trees, 535 or 90.98% were successfully detected by FSCT. The mean and median errors are within the expected error of the reference diameter measurements, suggesting minimal bias.”
Taking into account the importance of the FSCT tool used, it would be advisable to comment on the methodology used to make comparisons.
Comments are included in the Material and Methods section.
- g) A more in-depth analysis of the errors obtained in Table 3 (L267) is needed. In plots with conditions
On the other hand, it is necessary to indicate that it has not been possible to verify the correct citation of the bibliographic references as they appear as errors in the text (Error! Reference source not found)
Thanks for the notice, the errors have been removed.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsI agree with the author's revision, and I currently do not have any new comments.
Reviewer 4 Report
Comments and Suggestions for AuthorsFirst of all, I would like to thank the authors for the work carried out in this second review.
Undoubtedly, I believe that the work has gained clarity of exposition, however, the work still presents problems in the design of the experiments that greatly limit the quality of the work and the interest of its readers.
In this respect, if the objective is to compare the use of SLAM systems with classical techniques, it is important to carry out a complete analysis of the factors that can affect the errors inherent to the new systems (SLAM). In the work, it is proposed from the point of view of analysis of results, in some cases, with doubtful relationships between the errors detected and different variables that lead to conclusions that I am not clear whether they can be extrapolated to other case studies.
Basic aspects in this type of study, such as the extension and orography of the surface to be controlled, the trajectories followed by the SLAM system itself -the distance between passes and/or use of closed routes-, among others, are not considered in the study. This is something that I already pointed out in my first review, knowing the complexity of modifying the timeframe proposed for the review. It is important to point out that the authors acknowledge the interest of the comments but postpone the inclusion of these aspects for a future article.
In this sense, I consider that the most appropriate decision is to suggest to the authors that they carry out these modifications and, once the variables of interest have been correctly analyzed, prepare the corresponding manuscript with a view to its publication in this journal.
The key should be to provide guidelines for those readers interested in applying these methodologies in their professional work, not to limit themselves to presenting concrete results in a scheme of work that should be considered as difficult and difficult to understand.
