Research on Railway Obstacle Detection Method Based on Developed Euclidean Clustering

Round 1

Reviewer 1 Report

The topic is of interest and it is worth the research effort. It is challenging to find a solution for this problem. The paper presents a development but it does not solve the problem. It rather confirms that more effort is required on this front. 

To be addeqautely positioned in the current state of the art this paper should include a section on State of the art. Such a section is currently missing. The jargon used is not for a wide audience and needs to explained properly. The three methods mentioned should be introduced followed by a discussion on their strengths and weaknesses. At the moment this is not visible in the paper, as such the outreach of this publication will be quite limited. All abbreviations should be meaningful and clear to the wide audience to open up opportunities for intake elsewhere. Not only locally. 

The conclusions should be strengthened. e.g. Try to explain what is in your 1st dataset without mentioning it.

What future work do you envisage to be undertaken as a result of this research development?  

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript seems original, authors' own work, with a low similarity of 13% when checked in anti-plagiarism software.

Please allow me a few recommendations to improve the final aspect of the manuscript.

Is it lidar or LIDAR with capital letters? (the authors write this acronym differently even in the abstract). I think the term should be explained a little, at least for the readers of this journal.

I do not understand very clearly the expression in the abstract, does this phrase contain instructions or  is an explanation of the actions of the authors of this manuscript?: "Use the Random Sample Consensus (RANSAC) algorithm to remove the subgrade plane, fit the point cloud data on the rail, and divide the detection area based on the fitting findings."

I recommend a major check of the English language, possibly by a native speaker. It also seems a mixture of tenses (past and present) in English, starting from the abstract.

I think the researchers should explained more seriously why this particular RANSAC algorithm was chosen for the experiment carried out.

It is quite old, with a history behind it, and it also has a series of shortcomings, such as that there is no upper bound on the time it takes to compute the parameters, requires the setting of problem-specific thresholds, etc.

Perhaps another solution would have been Hough transform, an algorithm that is well known for its ability to detect straight line segments in the presence of noise, or other algorithms (or perhaps a comparison between some).

The list of bibliographic references seems rather/very short . I believe that the authors should increase their number with other innovative studies for the studied field, to which to report and with which to compare the results of their research. It would be interesting to emphasize more clearly towards the end of the manuscript (in a section/sub-section called Discussions, perhaps), what the authors bring new compared to other studies, what is their scientific contribution to the studied field.

I think that the technologies and methods used should be more detailed. Otherwise, it seems like a current, novel experiment, with a few improvements the manuscript can be published.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf