Sustainable Road Pothole Detection: A Crowdsourcing Based Multi-Sensors Fusion Approach
, Gaoang Wang 1, Chao Wu 5 and Simon Hu 1,2,6,*Round 1
Reviewer 1 Report
Please see the attachment.
Comments for author File: 
Comments.pdf
Author Response
Please see the file attached!
Author Response File: Author Response.pdf
Reviewer 2 Report
This manuscript proposes a road anomaly detection system based on multi-source sensor fusion (Accelerometer and image analysis).
The study is well structured and the experimental investigation is articulated and accurate.
The results and validations are also excellent and clearly express the thesis put forward by the authors.
In the literature analysis, the authors should better mention the use of accelerometric data from smartphones to monitor lateral friction and the consequent behavior of the driver.
This study is noted:
Vaiana R. et Al. (2017). Demanded vs Assumed friction along horizontal curves: an on-the-road experimental investigation. JOURNAL OF TRANSPORTATION SAFETY & SECURITY, p. 1-27, ISSN: 1943-9962, doi: 10.1080/19439962.2016.1277290
Author Response
Please see the file attached!
Author Response File: Author Response.pdf
Reviewer 3 Report
This manuscript presented a pothole detection strategy based on processing the cell phone accelerometer, GPS and the video image with a joint of multiple machine learning algorithms. The paper is well written with sufficient background information and clear presentation on the methodology and result. The reviewer suggests publication with minor revisions.
1. Line 60: Please cite the mentioned machine learning method that are small scale.
2. Line 63: Please cite “other studies.”
3. Line 72: Please cite “A significant number of studies.”
4. Line 97-113. Please cite the relevant paper regarding the “LSTM” “OPTICS” “DBSCAN”, IF they are not originally developed in this manuscript. Ignore these comments if these method are developed in this manuscript.
5. Line 195: “Xiao Li et al.[25]” There is no author names Xiao in reference 25.
6. Line 350 “Taspinar” Add relevant citation.
7. Line 500 Cite the relevant paper of “DBSCAN” if it is not developed in this manuscript.
8. Did the author tested the machine learning model on a section of road that never included in the training data set? Ideally, the verification test need to use the data that are not used in training models in order to truly justify the validation of the method. If the author has done that, please emphasis to take the credit. Otherwise, it is OK to not mentioning it.
Author Response
Please see the file attached!
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The author has done a good job with the revisions. In addition, I have a few remaining further questions/remarks.
1. In Introduction: This manuscript should further add some articles and will be of interest to many, such as:
a) Automatic Detection of Pothole Distress in Asphalt Pavement Using Improved Convolutional Neural Networks, https://doi.org/10.3390/rs14163892
b) Real-Time Road Pothole Mapping Based on Vibration Analysis in Smart City, https://doi.org/ 10.1109/JSTARS.2022.3200147
c) Road damage detection using UAV images based on multi-level attention mechanism, https://doi.org/10.1016/j.autcon.2022.104613
2. Please check the picture quality again to meet the publication requirements.
Author Response
Point 1: 1. In Introduction: This manuscript should further add some articles and will be of interest to many, such as:
- a) Automatic Detection of Pothole Distress in Asphalt Pavement Using Improved Convolutional Neural Networks, https://doi.org/10.3390/rs14163892
- b) Real-Time Road Pothole Mapping Based on Vibration Analysis in Smart City, https://doi.org/ 10.1109/JSTARS.2022.3200147
- c) Road damage detection using UAV images based on multi-level attention mechanism, https://doi.org/10.1016/j.autcon.2022.104613
Response 1: Thanks for your insightful comment and the recommended articles. We have reviewed them carefully, they are indeed quite relevant. We have now incoporated them to our literature review. They can be found in the text at lines 234 to 241 and lines 253 to 255.
Point 2: Please check the picture quality again to meet the publication requirements.
Response 2: Thanks for your careful comment. We have checked the picture quality and they have met the publication requirements.
