Design and Implementation of a Highly Scalable, Low-Cost Distributed Traffic Violation Enforcement System in Phuket, Thailand

Round 1

Reviewer 1 Report

The paper proposed a new traffic violation enforcement system that presents some advantages over other ones presented in the literature.

In general, the paper is very interesting; is well-written and easy to follow.

However, some changes need to be made before publication.

• Firstly, the first paragraph of the Introduction, where authors explain their contribution, should be extended more. Now, it is shorter than the first paragraph, where the authors explain the problem of traffic accidents.
• Before explaining the system Architecture (Section 2), a section describing the related work or state-of-art should be added.
• In the last paragraph of the Introduction, "We then give" should be replaced by "Then, we give".
• At the beginning of Section 2, "Next os the processor Layer". This sentence is not necessary. The same with the first sentence of Section 3.
• Authors should justify better the use of the two detection algorithms employed: Haar Adaboost cascade classifier and Convolutional Neural Networks, as well as the use of the Hog feature extractor.
• There is a mysterious blank space on page 10. Moreover, the content of such a page does not correspond with the content of page 9 but with the content of page 13.
• The deployment should not only be checked in intersections (Lines 284-285) because many traffic accidents happen in no intersections.
• The content of page 13 is followed on page 10, which is not logical.
• Finally, the conclusions section (Section 6) should be divided into more paragraphs.

Author Response

We received thoughtful peer reviews from three experts. We replicate their comments and explain how we have addressed each concern below. Reviewer comments are replicated in filled top-level bullet points, and our responses are provided in the corresponding sub-bullet points.   ===== Reviewer 1 =====

• Firstly, the first paragraph of the Introduction, where authors explain their contribution, should be extended more. Now, it is shorter than the first paragraph, where the authors explain the problem of traffic accidents.
  • The contributions described in the Introduction (paragraphs 3-4) have been revised and expanded.

• Before explaining the system Architecture (Section 2), a section describing the related work or state-of-art should be added.
  
  • A new section (Related Work) has been added. This section describes the methods and algorithms adopted in similar work and our system implementation.

• In the last paragraph of the Introduction, "We then give" should be replaced by "Then, we give".
  • The paragraph is edited as mentioned, and the manuscript has been thoroughly revised for English grammar and usage.

• At the beginning of Section 2, "Next os the processor Layer". This sentence is not necessary. The same with the first sentence of Section 3.
  • They are edited as mentioned.

• Authors should justify better the use of the two detection algorithms employed: Haar Adaboost cascade classifier and Convolutional Neural Networks, as well as the use of the Hog feature extractor.
  • We discuss the tradeoffs between classical methods and CNNs in Section 2 (Related Work).

• There is a mysterious blank space on page 10. Moreover, the content of such a page does not correspond with the content of page 9 but with the content of page 13. The content of page 13 is followed on page 10, which is not logical.
  • We apologize for the incorrect formatting, but this does not occur when we generate the PDF locally -- it seems to be introduced by the MDPI LaTeX compilation process. In case this error persists, we are attaching a correct PDF file generated locally with this reply.

• The deployment should not only be checked in intersections (Lines 284-285) because many traffic accidents happen in no intersections.
  • The reason for this decision by the policy makers is added to the 4th paragraph of the Introduction section.

• Finally, the conclusions section (Section 6) should be divided into more paragraphs.
  • We have edited according to the suggestion.

We are grateful to Reviewer 1 for the careful attention to our manuscript.

Author Response File: Author Response.pdf

Reviewer 2 Report

It is very interesting well-written paper. 

Author Response

We are grateful to the reviewer for the kind words.

We apologize for the incorrect formatting in the previous submission, but this does not occur when we generate the PDF locally -- it seems to be introduced by the MDPI LaTeX compilation process. In case this error persists, we are attaching a correct PDF file generated locally with this reply.

Author Response File: Author Response.pdf

Reviewer 3 Report

This article has proposed a new structure to monitor motorists’ compliance rate in Thailand putting forward new techniques for detecting the vehicles who fail to abide by traffic laws. Drawing upon some existing articles in this field, the authors have come up with a more cost-effective approach leading to improved safety and compliance rate among road users.  

The paper defines the problem appropriately and outlines the steps taken to enhance the existing literature at an acceptable level. However, authors should make some minor changes to the paper before it is published online.

1. There are some small language errors across the paper that need attention, such as line 216 ‘prove’.
2. After line 187, it has jumped into line 295. This gap must be addressed.
3. Figures 8 to 10 should be presented in the relevant section and page rather than pushing them to the end of the paper.

Author Response

We replicate their comments and explain how we have addressed each concern below. Reviewer comments are replicated in filled top-level bullet points, and our responses are provided in the corresponding sub-bullet points.

• There are some small language errors across the paper that need attention, such as line 216 ‘prove’.
  • We apologize for overlooking this and other typos and have thoroughly revised the paper for English grammar and style.
• After line 187, it has jumped into line 295. This gap must be addressed.
  • We apologize for the incorrect formatting, but this does not occur when we generate the PDF locally -- it seems to be introduced by the MDPI LaTeX compilation process. In case this error persists, we are attaching a correct PDF file generated locally with this reply.
• Figures 8 to 10 should be presented in the relevant section and page rather than pushing them to the end of the paper.
  • We appreciate the comment and would normally agree with the reviewer, but in this particular case, we felt that since the charts contain many lines and bars, we should make each of them and all corresponding labels clearly visible, requiring a full page. We also felt that spreading the content of these two pages of figures among the content of the article would interrupt the reader's continuity. So while our preference is to keep these pages as is, we will certainly follow the Editor's instructions if asked to make this change.

Author Response File: Author Response.pdf