تشخیص کلاه ایمنی موتورسواران توسط دوربین‌های ترافیکی در حالت‌های دشوار به کمک یادگیری عمیق

نوع مقاله : مقاله پژوهشی

نویسندگان

1 کارشناسی ارشد، دانشگاه صنعتی مالک اشتر، تهران، ایران

2 استادیار، دانشگاه علوم و فنون هوایی شهید ستاری، تهران، ایران

چکیده

کاهش مصدومیت و مرگ‌‌ومیر ناشی از حوادث رانندگی همواره موردتوجه مسئولین انتظامی و دولت‌ها می‌باشد. برای کاهش مصدومیت در مکان‌هایی که میزان وقوع حوادث رانندگی به علت عدم استفاده موتورسواران از کلاه ایمنی در آن زیاد می‌باشد اقدامات قابل‌توجه‌ای ازجمله حضور افسر انتظامی صورت گرفته است. تمامی این موارد توسط عوامل انسانی صورت گرفته که ممکن است مواردی از قبیل تعداد اندک کارکنان و خستگی آن‌ها کیفیت این نظارت را کاهش داده و نتیجه دلخواه را به همراه نداشته باشد. یکی از موارد تخلفات رانندگی عدم استفاده از کلاه ایمنی توسط موتور‌سواران است، راهکار ارائه‌شده در این پژوهش بهره‌بردن از الگوریتم‌های یادگیری عمیق جهت تشخیص استفاده یا عدم استفاده کلاه ایمنی توسط موتور‌سواران می‌باشد، سیستم پیشنهادی علاوه بر تشخیص می‌تواند‌ در تحلیل داده‌های ترافیکی ازجمله، در چه زمان‌هایی تخلفات کاهش یا افزایش می‌یاید مورداستفاده قرار گیرد. در این پژوهش جهت تشخیص کلاه ایمنی از سه نسخه 416، 320 و spp شبکه عصبی عمیق YOLO_v3 استفاده‌شده است و با عنایت به اینکه این شبکه‌ها از قبل بر روی دادگان COCO آموزش‌دیده‌اند از 53 لایه اول شبکه به‌صورت یادگیری انتقالی استفاده‌شده و 53 لایه دیگر بر اساس داده‎‌های مورداستفاده در این پژوهش مورد آموزش قرارگرفته و سپس عملکرد این سه شبکه با یکدیگر مقایسه شد. درنهایت سیستم پیشنهادی تشخیص خودکار کلاه ایمنی پس از آموزش توسط مجموعه داده Helmet Detection عملیات تشخیص را با مقدار mAP ، 08/96% انجام‌ می‌دهد، همچنین مدل با تعدادی از کارهای ‌پیشین مقایسه گردید و نتایج مطلوب‌تری حاصل شد.

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

Detection of motorcycle helmets by traffic cameras in difficult situations using deep learning

نویسندگان [English]

  • Masoud parvaneh 1
  • Pezhman Gholamnezhad 2
1 Master's degree, Malek Ashtar University of Technology, Tehran, Iran
2 Assistant Professor, Shahid Sattari University of Aeronautical Sciences and Technology, Tehran, Iran
چکیده [English]

Reducing injuries and deaths caused by traffic accidents is always the concern of law enforcement officials and governments. In order to reduce injuries in places where the rate of traffic accidents is high due to motorcycle riders not wearing helmets, significant measures have been taken, including the presence of police officers. All these cases are done by human factors, which may reduce the quality of this monitoring and may not bring the desired result, such as the small number of employees and their fatigue. One of the driving violations is the non-use of helmets by motorcyclists. The solution presented in this research is to use deep learning algorithms to detect the use or non-use of helmets by motorcyclists. In addition to detection, the proposed system can analyze traffic data, including When violations are reduced or increased are used. In this research, three versions 416, 320 and spp of the YOLO_v3 deep neural network have been used to detect helmets, and considering that these networks have already been trained on COCO data, the first 53 layers of the network have been used as transfer learning and 53 layers Next, it was trained based on the data used in this research, and then the performance of these three networks was compared with each other. Finally, the proposed automatic helmet detection system performs the detection operation with 96.08% accuracy after being trained by the Helmet Detection dataset. Also, the model was compared with a number of previous works and more favorable results were obtained.

کلیدواژه‌ها [English]

  • Deep learning
  • neural network
  • computer vision
  • helmet

Smiley face

 

[1].    R. G. R. a. G. M. Kh Tohidul Islam, "Recognition of Traffic Sign Based on Bag-of-Words and Artificial Neural Network," MDPI, 2017,DOI:10.3390/sym9080138.
[2].    who, "Global status report on road safety 2018," https://www.who.int/publications/i/item/9789241565684, 2018.
[3].    O. B. ,. S. C. Madhuchhanda Dasgupta, "Automated Helmet Detection for Multiple Motorcycle Riders using CNN," in 2019 IEEE Conference on Information and Communication Technology (CICT), 2019 , DOI:10.1109/CICT48419.2019.9066191.
[4].    M. A. ,. M. B. A. ,. T. M. ,. Z. M. S. A. A. Tasbeeha Waris, "CNN-Based Automatic Helmet Violation Detection of Motorcyclists for an Intelligent Transportation System," Hindawi, 2022, DOI:10.1155/2022/8246776.
[5].    E. Soltanikazemi, A. Aboah, E. Arthur and B. K. Hatuwal, "Fine-Tuning YOLOv5 with Genetic Algorithm For Helmet Violation Detection," arXiv, 2023.
[6].    V. C. K. A.-N. ,. P. A. L. A. O. A. Geoffery Agorku, "Real-Time Helmet Violation Detection Using YOLOv5 and Ensemble Learning," arXiv, 2023.
[7].    K. S. ,. R. H. Sherin Eliyas, "Helmet, Violation, Detection Using Deep Learning," European Journal of Molecular & Clinical Medicine, 2019.
[8].    M. N. D. G. Vignesh Raj A G, "Helmet Detection using Single Shot Detector (SSD)," IEEE, 2021, DOI:10.1109/ICESC51422.2021.9532985.
[9].    A. A. T. Tanupriya Choudhury, "A Deep Learning Approach to Helmet Detection for Road Safety," NISCAIR, 2020, DOI:10.56042/jsir.v79i06.39579.
[10]. A. H. Sutikno, "Improving Detection Performance of Helmetless Motorcyclists Using the Combination of HOG, HOP, and LDB Descriptors," INASS, p. 428, 2021, DOI:10.22266/ijies2022.0228.39.
[11]. N. R. A. G. A. B. K. S. R. Kurkute, "IOT Based Smart System for the Helmet Detection," in International Conference on Sustainable Computing in Science, Technology & Management (SUSCOM-2019), 2019, DOI:10.2139/ssrn.3356793.
[12]. P. Z.-T. X. W. Zhong-Qiu Zhao, "Object Detection with Deep Learning: A Review," IEEE, 2019,DOI:10.1109/TNNLS.2018.2876865.
[13]. M. M. S. B. L. B. R. G. Tsung-YiLin, "Microsoft COCO: Common Objects in Context," arxiv, 2014.
[14]. W. D. R. S. L.-J. L. K. L. L. F.-F. Jia Deng, "ImageNet: A large-scale hierarchical image database," IEEE, 2009,DOI:10.1109/CVPR.2009.5206848.
[15]. K. D. K. R. R. J. D. A. M. B. Aditya Lohia, Bibliometric Analysis of One-stage and Two-stage Object Detection, 2021.
[16]. R. Z. W. Lixuan Du, "Overview of two-stage object detection algorithms," ICSP 2020, 2020,DOI:10.1088/1742-6596/1544/1/012033.
[17]. R. S. C. Hang Zhang, "Review on One-Stage Object Detection Based on Deep Learning," EAI.EU, 2022,DOI:10.4108/eai.9-6-2022.174181.
[18]. H. R. A. K. R. Mohammadreza Iman, "A Review of Deep Transfer Learning and Recent Advancements," MDPI, 2023,DOI:10.3390/technologies11020040.
[19]. pytorch, "FINETUNING TORCHVISION MODELS," https://pytorch.org/tutorials/beginner/finetuning_torchvision_models_tutorial.html, 2023.
[20]. S. D. G. F. Joseph Redmon, "You Only Look Once:Unified, Real-Time Object Detection," IEEE, 2016، DOI:10.1109/CVPR.2016.91.
[21]. "YOLO: Real-Time Object Detection," https://pjreddie.com/darknet/yolo/, 2023.
[22]. W. L. L. ,. L. YUAN DAI, "Efficient Foreign Object Detection between PSDs and Metro Doors via Deep Neural Networks," IEEE, 2020, DOI:10.1109/ACCESS.2020.2978912.
[23]. "Helmet Detection," https://www.kaggle.com/datasets/andrewmvd/helmet-detection, 2023.
[24]. https://github.com/heartexlabs/labelImg, 2023.
[25]. S. L. N. A. B. d. S. Rafael Padilla, "A Survey on Performance Metrics for Object-Detection Algorithms," IEEE, 2020,DOI:10.1109/IWSSIP48289.2020.
[26]. "Common objects in Context," https://cocodataset.org/#detection-eval, 2023.
[27]. "Intersection over Union (IoU) for object detection," https://pyimagesearch.com/2016/11/07/intersection-over-union-iou-for-object-detection/, 2024.
[28]. "Confusion Matrix," https://wiki.cloudfactory.com/docs/mp-wiki/metrics/confusion-matrix, 2024.
[29]. "Recall score," 2024, https://wiki.cloudfactory.com/docs/mp-wiki/metrics/recall.
[30]. "Precision score," https://wiki.cloudfactory.com/docs/mp-wiki/metrics/precision, 2024.
[31]. "Average Precision," https://hasty.ai/docs/mp-wiki/metrics/average-precision, 2023.
[32]. "Mean Average Precision (mAP) Explained: Everything You Need to Know," https://www.v7labs.com/blog/mean-average-precision, 2023.
[33]. G. Colab, ""Google Colab"," https://colab.research.google.com/notebooks/intro.ipynb?utm_source=scs-index, 2023.
[34]. X. Z. S. R. a. J. S. Kaiming He, "Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition," IEEE, 2015، DOI:10.1109/TPAMI.2015.2389824.
[35]. "Real-time Multi-Class Helmet Violation Detection Using Few-Shot Data Sampling Technique and YOLOv8," in Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2023.
[36]. A. D. ,. B. ,. I. E. T. Elham Soltanikazemi, "Real-Time Helmet Violation Detection in AI City Challenge 2023 with Genetic Algorithm-Enhanced YOLOv5," in 2023 IEEE Applied Imagery Pattern Recognition Workshop (AIPR), 2023.