3D Landslide Monitoring in High Spatial Resolution by Feature Tracking and Histogram Analyses Using Laser Scanners

Round 1

Reviewer 1 Report

Comments on remotesensing-2725257-peer-review-v1

Title:  3D landslide monitoring in high spatial resolution by feature 2 tracking and histogram analyses using laser scans

The paper focuses on improving high spatial resolution landslide detection, a critical aspect of mitigating the impacts of landslides. It introduces an approach that involves using histogram analysis on displacement vectors to remove outliers generated during the matching algorithm. The research demonstrates the effectiveness of this method, showing that it maintains the accuracy and distribution of displacement vectors while reducing computational complexity. The paper emphasizes the significance of managing outliers in landslide prediction, providing a valuable contribution to landslide research and mitigation strategies. However, some general comments are as follows and followed by other concerns given below.

1.      Improve grammar and punctuation.

2.      Connection between paragraphs and sentences makes reading hard. Improve it.

3.      The structure and organization of the paper need major recommendations. For example, see my comments on Fig. 4. See the comments on the pdf file.

Abstract

1.      The abstract provides a clear overview of the problem addressed and the proposed solution involving histogram analysis for outlier removal. However, it could benefit from explicitly stating the novelty or unique contribution of the research. What sets this method apart from existing approaches? This additional information will help readers quickly grasp the significance of the research.

2.      In addition, while the abstract mentions that the proposed method has been tested on three different datasets, it would be beneficial to include some quantitative results to reinforce the effectiveness of the approach. Providing specific metrics or statistics on the performance improvement (e.g., percentage reduction in outliers, computational time saved) would enhance the credibility of the presented method. This addition would give readers a more concrete understanding of the research outcomes.

Overall, the abstract requires to be re-written.

Introduction:

1.      The introduction provides a comprehensive overview of landslides, their impacts, and the importance of studying active slope displacements. However, consider breaking down the introduction into smaller paragraphs to enhance readability. This will make it easier for the reader to follow the information and grasp key points.

2.      Consider providing specific citations or sources when discussing statistics or referencing organizations like the World Health Organization. This adds credibility to the information presented. For example, mentioning the specific report or publication from the WHO and including the publication date would strengthen the argument.

3.      Towards the end of the introduction, there is a shift towards discussing high-spatial-resolution monitoring methods. Consider adding a sentence or two to smoothly transition into this topic. This will help readers understand the logical progression of the paper.

Review literature:

1.      Consider breaking down the literature review into subsections for each method (GNSS, Image-based monitoring, Terrestrial Laser Scanners). This will provide a clearer and more structured overview for the reader, making it easier to digest the information.

2.      While the review provides a detailed explanation of each method, consider adding a brief summary or conclusion at the end of each subsection. This will help reinforce the key points and highlight the significance of each method in landslide monitoring.

3.      Ensure to integrate specific citations within the text to support the provided information. For example, when mentioning statistics or specific studies, include the corresponding citations to add credibility to the information.

Study area:

1.       Provide a bit more context about why each data set was chosen and its relevance to the study. For example, why was the simulated laboratory dataset used for initial testing, and what were the specific objectives for this dataset?

2.      Maintain consistent terminology and units when describing measurements and distances. For instance, in the first dataset, you mention measurements in meters, but in the third dataset, you use x-axis and y-axis without specifying units.

3.      It is good that you mention the georeferencing process, but consider providing a bit more detail on how it was done for each dataset. What specific points were used, and why were they chosen? This could provide valuable insight into the accuracy and reliability of the georeferencing.

4.      Shift most highlighted part of this section. Create another section.

Method:

1.      While the methodology is detailed, consider breaking down the steps into subsections for improved readability. This will help the reader navigate through the process more easily.

2.      When discussing the threshold value for outlier removal (e.g., 20 cm), providing a brief justification or rationale for choosing this specific value would be beneficial. For instance, you mentioned the expected movement rates of the object but briefly explained why this information influenced the threshold selection.

3.      Including diagrams or figures to illustrate the process could greatly enhance understanding. For example, having visual representations of the histograms and how they help identify outliers would provide a clearer grasp of the methodology.

4.      It might be helpful to briefly mention any potential limitations or challenges associated with this methodology. This could include factors like sensitivity to specific types of terrain or potential sources of error.

Results:

1.      The explanations are clear and detailed, making it easy to follow the evaluation process. However, in the section "Matching process and histogram analyses," you could provide a brief overview of what KAZE and SIFT algorithms are for readers who may not be familiar with them.

2.      Consider adding figures to support the points made in this section. For example, visual representations of hillshades with different resolutions (as mentioned in section 4.1) or histograms showing outlier removal (as mentioned in section 4.2) would be beneficial.

3.      In section 4.2, after discussing the removal of outliers, it might be useful to briefly explain what these outliers represent. This could provide context for readers who may be unfamiliar with the term.

4.      When discussing the accuracy assessment in section 4.3, consider including metrics like percentage accuracy or error rates to provide a clearer understanding of the performance of the proposed method.

5.      It is great that you highlight the achievement of goals in your research. Providing a brief recap of the goals at the beginning of the results section could be helpful for readers to understand how each evaluation contributes to the overall success of the method.

Discussion:

1.      When discussing the impact of various factors (e.g., point cloud density, georeferencing, matching process) on accuracy, it would be helpful to provide some quantitative measures or examples to illustrate this impact. For instance, you could mention specific cases or experiments where changes in these factors led to noticeable differences in accuracy.

2.      In section 2, where you mention that thresholds in histogram analysis should be defined depending on the area, it would be beneficial to provide some practical guidance on how to determine these thresholds. This could include discussing factors that should be taken into account and providing examples or ranges for different scenarios.

3.      While you mention applying the process to three different datasets, consider discussing the generalizability of your method to a broader range of datasets. Are there specific types of datasets or conditions for which your method may be more or less effective? Providing insights on this aspect can help readers understand the potential applicability of your approach.

4.      It might be beneficial to briefly acknowledge any potential limitations or challenges associated with your approach. This could include factors that may affect the method's performance in certain scenarios and how these challenges could be addressed in future research.

5.      If available, consider including visual aids (e.g., graphs, diagrams) to support your discussion. Visuals can help illustrate key points and make complex concepts easier to understand.

Conclusion:

1.      In the conclusion, you mention the potential weaknesses of the method and suggest improvements. While this is a valuable insight, consider explicitly outlining specific future directions or areas of research that could address these limitations. This could include discussing potential technologies or methodologies that could enhance the method.

2.      When you mention that the method could be adapted for predicting landslides in the future, it would be helpful to briefly touch upon some of the practical challenges or considerations in implementing real-time prediction systems. This could involve factors like data collection infrastructure, computational requirements, and integration with existing monitoring systems.

3.      Consider discussing how scalable this method is for different geographical or geological contexts. Are there specific conditions or types of landslides where this approach may be particularly effective or face challenges? Providing insights into the method's applicability in different scenarios can help guide future research or applications.

4.      While you mention achieving trustworthy and accurate results, consider providing some specific quantitative measures or statistics to support this claim. For example, you could mention the range of accuracy achieved or compare the method's performance to existing approaches if relevant.

5.      Elaborate on the potential real-world impact of this research. How might the improved understanding of landslides and the developed monitoring method benefit communities, authorities, or organizations involved in landslide mitigation and response efforts? Providing a clear picture of the practical implications can underscore the significance of your work.

Comments for author File: Comments.pdf

Improve grammar and punctuation.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper "3D landslide monitoring in high spatial resolution by feature tracking and histogram analyses using laser scans" presents a significant contribution to geospatial analysis and landslide monitoring. The paper is within the scope of the journal. and its special issue

The abstract outlines the primary challenge in feature-based landslide detection—namely, the prevalence of outliers in the matching algorithm, leading to computational inefficiency. The novel approach of using histogram analysis to identify and eliminate these outliers is a promising development that seems to enhance landslide monitoring accuracy without compromising the data quality.

The methodology proposed appears to be robust, as evidenced by its application to three different datasets. The use of terrestrial laser scanners for data acquisition is a strong choice due to their precision and reliability in capturing detailed surface information. However, the abstract lacks specific details on the feature tracking method used, which would be crucial for understanding the full scope of the research.

The Literature review is OK. It can be made more comprehensive by including the latest studies. For eg : Image based methods lack mention of UAV-based GeoOBIA (You can see https://www.spiedigitallibrary.org/conference-proceedings-of-spie/12327/2666770/GeoBIA-based-semi-automated-landslide-detection-using-UAS-data/10.1117/12.2666770.short?SSO=1)

The conclusion reaffirms the effectiveness of the method, highlighting the potential for improved landslide monitoring strategies. The researchers' acknowledgment of the limitation regarding feature extraction in areas with smooth transitions demonstrates a good understanding of the method's constraints and an area for further research. The suggestion to incorporate aerial and satellite imagery to overcome this limitation is well-justified. It shows a pathway for integrating this method with other remote sensing technologies for enhanced feature extraction.

The future directions mentioned, such as real-time data collection and analysis for predictive capabilities, are particularly compelling. This advancement could be revolutionary for geohazard mitigation, offering a proactive rather than reactive approach to landslide management.

Figure 3 can be improved. The label in 3 a makes it look quite clumsy.

The discussion section is very short.

However, the paper would benefit from a more detailed discussion on the following:

Algorithm Efficiency: While the histogram analysis helps reduce outliers, the efficiency of the algorithm could be further scrutinized, especially in terms of its computational time and resources.

Validation Techniques: More information on how the accuracy was assessed against ground truth or through field validation would be beneficial.

Scalability: The application to larger, more complex landslide events should be discussed, along with the method's scalability and limitations in different environments and conditions.

Comparative Analysis: A comparison with existing methods and technologies would provide a clearer picture of the proposed method's advantages and competitiveness in the field.

Feature Extraction Details: Expanding on the feature extraction process used, including the types of features and the matching algorithm, would provide greater insight into the method's robustness.

Integration with Predictive Models: The paper hints at future predictive capabilities but does not detail how the method could be integrated with existing landslide prediction models.

Societal Impact: A discussion on how this technology could be practically implemented for community early-warning systems would be beneficial.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

authors have nearly incorporated all the recommendations. I accept all the changes done by the authors. now the quality of the manuscript is quite good. But authors need to justify why KAZE was used and why not AKAZE or any improved version of AKAZE was selected. Also, use high-resolution figure for figure 3. One which is at 300 DPI would be good.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf