Recent Research Progress on Ground-to-Air Vision-Based Anti-UAV Detection and Tracking Methodologies: A Review
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis contribution provides a comprehensive review of recent advancements in ground-to-air vision-based Anti-UAV detection and tracking systems. The authors discuss state-of-the-art object detection frameworks, including one-stage and two-stage detectors, and lightweight architectures for edge computing. The study also explores UAV tracking methodologies and evaluates publicly available datasets for anti-UAV research. Challenges and areas for future development are also discussed by the authors.
I suggest to improve the content to provide performance comparisons of the reviewed algorithms and/or add case studies or experimental results to validate the methods. In particular, the discussion on hybrid systems that integrate vision-based methods with radar or RF technologies might be extended to support the reader. In conclusion, this paper has the potential to make a valuable contribution to the field and can be accepted for publication with minor revisions.
Author Response
Thank you for your thoughtful feedback on our manuscript. We appreciate your constructive suggestions for improving the paper's contribution to the field. Please find the detailed responses below and the corresponding revisions in the attached file.
Comment 1: "I suggest to improve the content to provide performance comparisons of the reviewed algorithms and/or add case studies or experimental results to validate the methods."
Response 1: We added Table A1 in the Appendix Section and Table 2 on Page 13 that provide a performance comparison of the reviewed algorithms for detection and tracking. Additionally, we elaborated the Evaluation metrics used to make the performance evaluation of the discussed algorithms on Page 9 line 343 and Page 12 line 490.
Comment 2: “the discussion on hybrid systems that integrate vision-based methods with radar or RF technologies might be extended to support the reader”
Response 2: To address this point, we want to mention that Hybrid Anti-UAV systems which utilize vision-based methods and RF technologies are under-developed and as this is a review paper, we have decided to not delve too deep into the topic and leave it as a possible future research direction.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe paper reviews the research progress of ground-to-air vision-based anti-UAV detection and tracking methods, including algorithms, datasets, and systems. It also analyzes their advantages, disadvantages, and application scenarios, aiming to provide researchers with a comprehensive understanding, identify technological gaps, and point out directions for future research. However, some issues still need to be improved:
(1) The article mentions multiple publicly available data sets, but does not discuss in depth the coverage of these data sets under different environmental conditions (e. g., different weather, light conditions). Suggest that the authors are able to complement the analysis of the representativeness and limitations of the dataset under these conditions.
(2) Although the article mentions the challenges of multi-drone tracking, the discussion of the progress and future directions of research in this area is not deep enough. It is suggested that the authors can discuss the current development status and major challenges of multi-UAV tracking technology in detail.
(3) The article mentions the integration with the UAV traffic management system (UTM), but lacks specific implementation strategies and case analysis. Suggest that the authors are able to provide some specific suggestions or case studies on how to integrate anti-UAV systems with existing UTM frameworks.
(4) It is recommended that the author can add charts and visualization elements to help readers more intuitive geosolution performance and data set features.
(5) There are many professional terms and abbreviations in the text, so how to ensure that readers who are not familiar with the field can easily understand them. Do you need to add more explanations or auxiliary instructions?
Comments on the Quality of English LanguageThe language expression in this article is generally good and can be further condensed.
Author Response
Thank you so much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions in the attached file.
Comment 1: The article mentions multiple publicly available data sets, but does not discuss in depth the coverage of these data sets under different environmental conditions (e. g., different weather, light conditions).
Response 1: We have added Table 4 to include further characteristics of the datasets, which strengthen the descriptions provided in Section 4.1.
Comment 2: Although the article mentions the challenges of multi-drone tracking, the discussion of the progress and future directions of research in this area is not deep enough. It is suggested that the authors can discuss the current development status and major challenges of multi-UAV tracking technology in detail.
Response 2: Thank you for pointing this out. To address this, we have improved the discussion on multi-UAV detection and Tracking in Section 5, line 647. It is to be noted that little to no work has been done to identify and track multiple rogue UAVs and separate them from a swarm of regular UAV traffic.
Thus, we are unable to discuss it in more detail in this review paper. The challenges of multi-UAV tracking are also discussed in Section 3.2.
Comment 3: The article mentions the integration with the UAV traffic management system (UTM), but lacks specific implementation strategies and case analysis. Suggest that the authors are able to provide some specific suggestions or case studies on how to integrate anti-UAV systems with existing UTM frameworks.
Response 3: To address this, we have improved the discussion on Integration into the UTM frameworks in Section 5, line 672.
Comment 4: It is recommended that the author can add charts and visualization elements to help readers more intuitive geosolution performance and data set features.
Response 4: We have added Table 2 and 4 to better visualize and intuitively inform the readers.
Comment 5: There are many professional terms and abbreviations in the text, so how to ensure that readers who are not familiar with the field can easily understand them.
Response 5: All abbreviations and terms are listed on Page 18 under Abbreviations. Terms that may require further understanding have relevant citations provided beside them.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript for this review paper has nicely covered aspects (number of stages, anchors, neck etc.) of how CNN-based architectures (with special emphasis on those related to YOLO family) play their respective roles in detection accuracy. However, as a review paper, a wider look at the problem at hand is needed. In the recent past, various researchers have put together multiple review papers on the same problem. For instance, please look at the following recent review article published in MDPI Drones:
Liu, Z.; An, P.; Yang, Y.; Qiu, S.; Liu, Q.; Xu, X. Vision-Based Drone Detection in Complex Environments: A Survey. Drones 2024, 8, 643. https://doi.org/10.3390/drones8110643
This review takes a look at the problem at hand from various angles e.g.
a) Data acquisition scenarios related to dim light, occlusion, backgrounds and not just listing of various available datasets
b) Artificial dataset generation and automatic labelling techniques
c) A comprehensive list of practical problems (Fig. 6)
d) InfraRed Imaging approaches
e) Evaluation metrics
Other relevant and recent review works also need to be cited since multiple such papers have been published and the current work has to convey more useful further information to be beneficial
a) Seidaliyeva, U.; Ilipbayeva, L.; Taissariyeva, K.; Smailov, N.; Matson, E.T. Advances and Challenges in Drone Detection and Classification Techniques: A State-of-the-Art Review. Sensors 2024, 24, 125. https://doi.org/10.3390/s24010125
b) Wang, B.; Li, Q.; Mao, Q.; Wang, J.; Chen, C.L.P.; Shangguan, A.; Zhang, H. A Survey on Vision-Based Anti Unmanned Aerial Vehicles Methods. Drones 2024, 8, 518. https://doi.org/10.3390/drones8090518
c) Nader Al-lQubaydhi, Abdulrahman Alenezi, Turki Alanazi, Abdulrahman Senyor, Naif Alanezi, Bandar Alotaibi, Munif Alotaibi, Abdul Razaque, Salim Hariri, Deep learning for unmanned aerial vehicles detection: A review, Computer Science Review, Volume 51, https://doi.org/10.1016/j.cosrev.2023.100614.
Author Response
Thank you so much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions in the attached file.
Comment 1: Other relevant and recent review works also need to be cited since multiple such papers have been published and the current work has to convey more useful further information to be beneficial.
Response 1: Thank you for pointing this out! We have cited the mentioned recent surveys to better explain how our manuscript differs from and contributes to this research domain. We have also included a better explanation of the intuition behind the motivation for this review paper in Section 1, Line 51. Table 2 (similar to Table A1) has also been added to further enhance the value of this review from a ground-to-air Anti-UAV tracking perspective.
Comment 2: Data acquisition scenarios related to dim light, occlusion, backgrounds and not just listing of various available datasets.
Response 2: Section 4.1 discusses the pros, cons, and characteristics of the listed datasets in Table 3. We added Table 4 to include another summarized perspective.
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
Comments and Suggestions for AuthorsThe revised manuscript has adequately addressed all the concerns raised in the previous review cycl.e
