A Comparative Assessment of Deep Neural Network Models for Detecting Obstacles in the Real Time Aerial Railway Track Images

References

• Indian Railways Annual Report 2019-2020. 2020. Available online: https://indianrailways.gov.in/railwayboard/uploads/directorate/stat_econ/Annual-Reports-2019-2020/Indian-Railways-Annual%20-Report-Accounts%20-2019-20-English.pdf (accessed February 10, 2021).
   Google Scholar
• Machine learning crash course - Google’s best practices on splitting data. 2020. Available online: https://tinyurl.com/y7yqfhxu (accessed February 28, 2021).
   Google Scholar
• Indian Railways procures Ninja drones for security surveillance and passenger safety 2021. Available Online: https://economictimes.indiatimes.com/industry/transportation/railways/indian-railways-introduces-drone-based-surveillance-system-for-rail-security/articleshow/77621214.cms?from=mdr (accessed February 10, 2021).
   Google Scholar
• Akkas, S., S. Singh Maini, and J. Qiu. 2019. A fast video image detection using tensor flow mobile networks for racing cars. IEEE International Conference on Big Data, Los Angeles, CA, USA, doi: 10.1109/BigData47090.2019.9005689.
   Google Scholar
• Breton, M. 2019. Overview of two performance metrics for object detection algorithms evaluation. Defence Research and Development Canada. DRDC-RDDC-2019-D168.
   Google Scholar
• Chandran, P., F. Thierry, J. Odelius, S. M. Famurewa, H. Lind, and M. Rantatalo. 2021. Supervised Machine Learning Approach for Detecting Missing Clamps in Rail Fastening System from Differential Eddy Current Measurements. Applied Sciences, MDPI 11 (9):4018. doi:10.3390/app11094018.
   Google Scholar
• Chiu, Y.-C., C.-Y. Tsai, M.-D. Ruan, G.-Y. Shen, and T.-T. Lee. 2020. Mobilenet-SSDv2: an improved object detection model for embedded systems. International Conference on System Science and Engineering (ICSSE), Kagawa, Japan. doi: 10.1109/ICSSE50014.2020.9219319
   Google Scholar
• Drone Surveillance in Indian Railways. 2020. Available online: https://economictimes.indiatimes.com/topic/drones-for-surveillance (accessed February 10, 2021).
   Google Scholar
• Duan, K., S. Bai, L. Xie, H. Qi, Q. Huang, and Q. Tian. 2019. Centernet: Keypoint triplets for object detection. Proceedings of the IEEE International Conference on Computer Vision 2019:6569–1517. arXiv:1904.08189.
   Google Scholar
• Fayyaz, M. A. B., and C. Johnson. 2020. Object Detection at Level Crossing Using Deep Learning. Micromachines, MDPI 11 (12):1055–78. doi:10.3390/mi11121055.
   PubMed Web of Science ®Google Scholar
• Flammini, F., R. Naddei, C. Pragliola, and G. Smarra. 2016. Towards automated drone surveillance in railways: State-of-the-art and future directions. Lecture Notes in Computer Science, vol. 100016, 336–34. 17th International Conference on Advanced Concepts for Intelligent Vision Systems, ACIVS 2016; Lecce; Italy: Springer. doi:10.1007/978-3-319-48680-2_30.
   Google Scholar
• Gebauer, O., W. Pree, and B. Stadlmann. 2012. Autonomously driving trains on open tracks—concepts, system architecture and implementation aspects. Information Technology 54:266–79.
   Google Scholar
• Ghiasi, G., T. Y. Lin, and Q. V. Le. 2018. Dropblock: A regularization method for convolutional networks. Advances in Neural Information Processing Systems 10727–37. arXiv:1810.12890.
   Google Scholar
• He K, Zhang X, and Ren S, 2017. Deep Residual Learning for Image Recognition. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 12:770–778.
   Google Scholar
• Hodge, V. J., S. O. Keefe, M. Weeks, and A. Moulds. 2015. Wireless sensor networks for condition monitoring in the railway industry: A survey. IEEE Transactions on Intelligent Transportation Systems 16 (3):1088–106. doi:10.1109/TITS.2014.2366512.
   Web of Science ®Google Scholar
• Howard, A. G., M. Zhu, B. Chen, D. Kalenichenko, W. Wang, T. Weyand, and H. Andreetto Mand Adam 2017. Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv:1704.04861v1.
   Google Scholar
• Indian Railways Permananet Way Manual. 2020. Available Online: https://www.iricen.gov.in/iricen/Track_Manuals/NEWIRPWM.pdf (accessed February 10, 2021).
   Google Scholar
• Jiao, L., F. Zhang, F. Liu, S. Yang, L. Li, Z. Feng, and R. Qu. 2019. A survey of deep learning-based object detection. IEEE Access 7:128837–68. doi:10.1109/ACCESS.2019.2939201.
   Google Scholar
• Kapoor, R., R. Goel, and A. Sharma. 2018. Deep learning based object and railway track recognition using train mounted thermal imaging system. Journal of Computational and Theoretical Nanoscience 17:5062–71. doi:10.1166/jctn.2020.9342.
   Google Scholar
• Kaster, J., J. Patrick, and H. S. Clouse 2017. Convolutional neural networks on small unmanned aerial systems. IEEE National Aerospace and Electronics Conference (NAECON), Dayton, OH, USA. 149–54. doi: 10.1109/NAECON.2017.8268760.
   Google Scholar
• Kumar Sen, P., M. Bhiwapurkar, and S. P. Harsha. 2018. Analysis of causes of rail derailment in India and corrective measures, reliability and risk assessment in engineering. Proceedings of INCRS 2018, Singapore: Springer, 305–14. doi: 10.1007/978-981-15-3746-2_28
   Google Scholar
• Li, S., X. Zhao, and G. Zhou. 2019. Automatic pixel-level multiple damage detection of concrete structure using fully convolutional network, Computer-Aided Civil and Infrastructure Engineering. John Wiley & Sons, Inc. United States: Wiley, 616–34. doi:10.1111/mice.12433.
   Google Scholar
• Liu, D., Z. Lu, T. Cao, and T. Li. 2017. A real-time posture monitoring method for rail vehicle bodies based on machine vision. International Journal of Vehicle Mechanics and Mobility, Vehicle System Dynamics, Taylor and Francis 55 (6):853–74. doi:10.1080/00423114.2017.1284339.
   Web of Science ®Google Scholar
• Liu, W., D. Anguelov, D. Erhan, C. Szegedy, S. Reed, F. Cheng-Yang, and C. B. Alexander. 2016. SSD: Single shot multibox detector. European Conference on Computer Vision 2016 9905:21–37. doi:10.1007/978-3-319-46448-0_2.
   Google Scholar
• Mittal, P., R. Singh, and A. Sharma. 2020. Deep learning-based object detection in low-altitude UAV datasets: A survey. Image and Vision Computing, Elsevier 104. doi:10.1016/j.imavis.2020.104046.
   Google Scholar
• Möckel, S., F. Scherer, and P. F. Schuster 2003. Multi-sensor obstacle detection on railway tracks. Proceedings of the IEEE IV2003 Intelligent Vehicles Symposium, Columbus, OH, USA. 42–46. doi: 10.1109/IVS.2003.1212880.
   Google Scholar
• Newell, A., K. Yang, and J. Deng 2016. Stacked hourglass networks for human pose estimation. European Conference on Computer Vision 2016 Amsterdam, Netherlands. arXiv:1603.06937
   Google Scholar
• Oliveira, F., T. Eilam, P. Nagpurkar, C. Isci, M. Kalantar, W. Segmuller, and E. Snible. 2016. Delivering software with agility and quality in a cloud environment. IBM Journal of Research and Development 60 (2):10:1–10:11. doi:10.1147/JRD.2016.2517498.
   Web of Science ®Google Scholar
• Padmanabhan, A. 2017. Civilian Drones and Indian Regulatory Response. Carnegie India: Center for Policy Research.
   Google Scholar
• Puppala, A. J., and S. Sarat Chandra Congress. 2019. A holistic approach for visualization of transportation infrastructure assets using UAV-CRP technology. International Conference on Information technology in Geo-Engineering, Springer Series in Geomechanics and Geoengineering Guimaraes, Portugal, 3–17. doi: 10.1007/978-3-030-32029-4_1
   Google Scholar
• Railway worker Killed. 2020. Available Online: https://timesofindia.indiatimes.com/city/visakhapatnam/1-killed-as-boulders-fall-on-train-track/articleshow/75562777.cms (accessed May 06, 2020).
   Google Scholar
• Rampriya, R. S., Sabarinathan, and R. Suganya. 2021. RSNet: Rail semantic segmentation network for extracting aerial railroad images. Journal of Intelligent and Fuzzy Systems, IOS Press 1–18. doi:10.3233/JIFS-210349.
   Google Scholar
• RDSO Important OHE Parameters. 2003. Available online: https://rdso.indianrailways.gov.in/works/uploads/File/Important%20OHE%20Parameters(2).pdf (accessed September 15, 2021).
   Google Scholar
• Reddy Pailla, D., V. Kollerathu, and S. S. Chennamsetty, 2019. Object detection on aerial imagery using CenterNet. arXiv:1908.08244v1.
   Google Scholar
• Redmon, J., and A. Farhadi 2018. Yolov3: An incremental improvement arXiv PreprarXiv1804.02767.
   Google Scholar
• Ren, S., K. He, R. Girshick, and J. Sun. 2015. Faster RCNN: Towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence 39 (6):1137–49. doi:10.1109/TPAMI.2016.2577031.
   Web of Science ®Google Scholar
• Risti, C., D. Durrant, M. A. Haseeb, M. Franke, M. Banic, M. Simonovic, and D. Stamenkovic. 2020. Artificial intelligence for obstacle detection in railways: Project SMART and beyond. Communication and Information Science Book Series Springer 1279:44–55.
   Google Scholar
• Ristic-Durrant, D., M. Franke, and K. Michels. 2021. A review of vision-based on-board obstacle detection and distance estimation in railways. Sensors, MDPI 21 (10):3452. doi:10.3390/s21103452.
   PubMed Web of Science ®Google Scholar
• Sandler, M., A. Howard, M. Zhu, A. Zhmoginov, and L. C. Chen 2018. MobileNetV2: Inverted residuals and linear bottlenecks. IEEE Conference on Computer Vision and Pattern Recognition Salt Lake City, Utah, USA. 4510–20. arXiv:1801.04381.
   Google Scholar
• Sinha, D., and F. Feroz. 2016. Obstacle detection on railway tracks using vibration sensors and signal filtering using bayesian analysis. IEEE Sensors Journal 16 (3):642–49. doi:10.1109/JSEN.2015.2490247.
   Web of Science ®Google Scholar
• Soliman, A., and J. Terstriep. 2019. Keras Spatial: Extending deep learning frameworks for preprocessing and on-the-fly augmentation of geospatial data. GeoAI 2019: Proc of the 3rd ACM SIGSPATIAL Intnl Workshop on AI for Geographic Knowledge Discovery Chicago IL, USA. 69–76. doi: 10.1145/3356471.3365240
   Google Scholar
• Srivastava, S., S. Narayan, and S. Mitta. 2021. A survey of deep learning techniques for vehicle detection from UAV images. Journal of Systems Architecture, Elsevier 117:102152. doi:10.1016/j.sysarc.2021.102152.
   Web of Science ®Google Scholar
• Suharto, E., Suhartono, A. P. Widodo, and E. A. Sarwoko. 2020. The use of mobilenet v1 for identifying various types of freshwater fish. Journal of Physics: Conference Series, IOP Publishing. doi:10.1088/1742-6596/1524/1/012105.
   Google Scholar
• Tan, M., and Q. V. Le 2019. Efficientnet: Rethinking model scaling for convolutional neural networks. Proceedings of the 36th International Conference on Machine Learning, Long Beach, CA, USA. 6105–14. arXiv:1905.11946.
   Google Scholar
• Tan, M., R. Pang, and Q. V. Le. 2020. EfficientDet: Scalable and efficient object detection. arXiv:1911.09070v7.
   Google Scholar
• Tolani, M., R. K. S. Sunny, K. Shubham, and R. Kumar. 2017. Two-layer optimized railway monitoring system using Wi-Fi and ZigBee interfaced wireless sensor network. IEEE Sensors Journal 17 (7):2241–48. doi:10.1109/JSEN.2017.2658730.
   Web of Science ®Google Scholar
• Ukai, M., B. Tomoyuki, and N. N. Nozomi 2011. Obstacle detection on railway track by fusing radar and image sensor. Proceedings of the 9th World Congress on Railway Research (WCRR), Lille, France. 1–12.
   Google Scholar
• Ukai, M. 2004. Obstacle detection with a sequence of ultra-telephoto camera images. WIT Transactions on the Build Environment 74:10. doi:10.2495/CR041001.
   Google Scholar
• Uribe J A, Fonseca L, and Vargas J F. 2012. Video Based System for Railroad Collision Warning. Proceedings of the IEEE International Carnahan Conference on Security Technology (ICCST), Newton, MA, USA, 280–285.
   Google Scholar
• US Department of transportation. 2018. Unmanned aircraft system applications in international railroads.
   Google Scholar
• Xu, Y., C. Gao, L. Yuan, S. Tang, and G. Wei 2019. Real-time obstacle detection over rails using deep convolutional neural network. Proceedings of the IEEE Intelligent Transportation Systems Conference (ITSC), Auckland, New Zealand. 1007–12. doi: 10.1109/ITSC.2019.8917091
   Google Scholar
• Yundong, L., H. Dong, L. Hongguang, X. Zhang, B. Zhang, and Z. Xiao. 2020. Multi-block SSD based on small object detection for UAV railway scene surveillance. Chinese Journal of Aeronautics 33 (6):1747–55. doi:10.1016/j.cja.2020.02.024.
   Web of Science ®Google Scholar
• Zhao, Z. Q., P. Zheng, S. Xu, and X. Wu 2018. Object detection with deep learning: A review. 3212–22. arXiv1807.05511 11.
   Google Scholar
• Zhou, J., C. M. Vong, Q. Liu, and Z. Wang. 2019. Scale adaptive image cropping for UAV object detection. Neurocomputing, Elsevier 366:305–13. doi:10.1016/j.neucom.2019.07.073.
   Web of Science ®Google Scholar