Journal Home > Volume 4 , Issue 1

Infant drowning has occurred frequently in swimming pools recent years, which motivates the research on automatic real-time detection of the accident. Unlike youths or adults, swimming infants are small in terms of size and motion range, and unable to send out distress signals in emergencies, which exerts negative effects on the detection of drowning. Aiming at this problem, a new step is initialized towards detecting infant drowning automatically and efficiently based on video surveillance. Diverse live-scene videos of infant swimming and drowning are collected from a variety of natatoriums and labeled as datasets. A part of the datasets is downscaled or enlarged to enhance generalization ability of the model. On this basis, advantages of Faster R-CNN and a series of YOLOv5 models are specifically explored to enable fast and accurate detection of infant drowning in real-world. Supervised learning experiments are carried out, model test results show that mean Average Precision (mAP) of either Faster R-CNN or YOLOv5s of the series of YOLOv5 can be over 89%; the former can process merely 6 frames of videos per second with the precision of only 62.04%, while the latter can reach an average speed of 75 frames/s with the precision of about 86.6%. The YOLOv5s eventually stands out as an optimal model for detecting infant drowning in view of comprehensive performance, which is of great application value to reduce the accidents in swimming pools.


menu
Abstract
Full text
Outline
About this article

Automatic Real-Time Detection of Infant Drowning Using YOLOv5 and Faster R-CNN Models Based on Video Surveillance

Show Author's information Qianen He1Zhiqiang Mei1Huisheng Zhang1Xiuying Xu1( )
School of Physics and Information Engineering, Fuzhou University, Fuzhou 350108, China

Abstract

Infant drowning has occurred frequently in swimming pools recent years, which motivates the research on automatic real-time detection of the accident. Unlike youths or adults, swimming infants are small in terms of size and motion range, and unable to send out distress signals in emergencies, which exerts negative effects on the detection of drowning. Aiming at this problem, a new step is initialized towards detecting infant drowning automatically and efficiently based on video surveillance. Diverse live-scene videos of infant swimming and drowning are collected from a variety of natatoriums and labeled as datasets. A part of the datasets is downscaled or enlarged to enhance generalization ability of the model. On this basis, advantages of Faster R-CNN and a series of YOLOv5 models are specifically explored to enable fast and accurate detection of infant drowning in real-world. Supervised learning experiments are carried out, model test results show that mean Average Precision (mAP) of either Faster R-CNN or YOLOv5s of the series of YOLOv5 can be over 89%; the former can process merely 6 frames of videos per second with the precision of only 62.04%, while the latter can reach an average speed of 75 frames/s with the precision of about 86.6%. The YOLOv5s eventually stands out as an optimal model for detecting infant drowning in view of comprehensive performance, which is of great application value to reduce the accidents in swimming pools.

Keywords: supervised learning, video surveillance, infant drowning detection, YOLOv5, Faster R-CNN

References(25)

[1]

M. N. Dai, Y. Xi, W. Q. Yin, Z. M. Chen, Z. Q. Feng, and C. H. Tang, Incidence, mortality and trends of drowning among children aged 0–14 years in China, 1990–2019, Chin. J. School Health, vol. 43, no. 2, pp. 256–259&264, 2022.

[2]

F. Wang, Y. B. Ai, and W. D. Zhang, Detection of early dangerous state in deep water of indoor swimming pool based on surveillance video, Signal,Image and Video Processing, vol. 16, no. 1, pp. 29–37, 2022.

[3]
A. Kulkarni, K. Lakhani, and S. Lokhande, A sensor based low cost drowning detection system for human life safety, in Proc. 5th Int. Conf. Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO), Noida, India, 2016, pp. 301–306.
DOI
[4]
F. Dehbashi, N. Ahmed, M. Mehra, J. Wang, and O. Abari, SwimTrack: Drowning detection using RFID, in Proc. ACM SIGCOMM 2019 Conf. Posters and Demos, Beijing, China, 2019, pp. 161–162.
DOI
[5]
A. Jose and G. Udupa, Gantry robot system for preventing drowning accidents in swimming pools, Mater. Today: Proc.,
[6]
Y. Y. Wang, Video monitoring system design and implementation of the underwater of pool, (in Chinese), Master dissertation, Beijing University of Technology, Beijing, China, 2014.
[7]

J. B. Hou and B. G. Li, Swimming target detection and tracking technology in video image processing, Microprocessors and Microsystems, vol. 80, p. 103535, 2021.

[8]

A. Alotaibi, Automated and intelligent system for monitoring swimming pool safety based on the IoT and transfer learning, Electronics, vol. 9, no. 12, p. 2082, 2020.

[9]

A. Claesson, S. Schierbeck, J. Hollenberg, S. Forsberg, P. Nordberg, M. Ringh, M. Olausson, A. Jansson, and A. Nord, The use of drones and a machine-learning model for recognition of simulated drowning victims-A feasibility study, Resuscitation, vol. 156, pp. 196–201, 2020.

[10]
M. A. Hayat, G. T. Yang, A. Iqbal, A. Saleem, A. Hussain, and M. Mateen, The swimmers motion detection using improved VIBE algorithm, in Proc. Int. Conf. Robotics and Automation in Industry (ICRAI), Rawalpindi, Pakistan, 2019, pp. 1–6.
DOI
[11]
A. I. N. Alshbatat, S. Alhameli, S. Almazrouei, S. Alhameli, and W. Almara, Automated vision-based surveillance system to detect drowning incidents in swimming pools, in Proc. Advances in Science and Engineering Technology International Conf. (ASET), Dubai, United Arab Emirates, 2020, pp. 1–5.
DOI
[12]

F. Lei, H. Y. Zhu, F. F. Tang, and X. Y. Wang, Drowning behavior detection in swimming pool based on deep learning, Signal Image and Video Processing, vol. 16, no. 6, pp. 1683–1690, 2022.

[13]
P. Pavithra, S. Nandini, A. Nanthana, N. T. Aslam, and P. Praveen Kumar, Video based drowning detection system, in Proc. Int. Conf. Design Innovations for 3Cs Compute Communicate Control (ICDI3C), Bangalore, India, 2021, pp. 203–206.
DOI
[14]

Y. J. Cha, W. Choi, and O. Büyüköztürk, Deep learning-based crack damage detection using convolutional neural networks, Computer-Aided Civil and Infrastructure Engineering, vol. 32, no. 5, pp. 361–378, 2017.

[15]

Y. J. Cha, W. Choi, G. Suh, S. Mahmoudkhani, and O. Büyüköztürk, Autonomous structural visual inspection using region-based deep learning for detecting multiple damage types, Computer-Aided Civil and Infrastructure Engineering, vol. 33, no. 9, pp. 731–747, 2018.

[16]
Y. N. Gao, Research on the method of drowning monitoring in infant swimming pool, (in Chinese), Master dissertation, Beijing University of Technology, Beijing, China, 2020.
[17]
Y. Deng and T. H. Zhou, Sensor-based self-rescue alarm system for the prevention of child drowning, in Proc. 13th Int. Conf. Intelligent Human-Machine Systems and Cybernetics (IHMSC), Hangzhou, China, 2021, pp. 168–171.
DOI
[18]
Y. Nishida, K. Hiratsuka, and H. Mizoguchi, Prototype of infant drowning prevention system at home with wireless accelerometer, in Proc. SENSORS, Atlanta, GA, USA, 2007, pp. 1209–1212.
DOI
[19]
K. Hiratsuka, Y. Nishida, and H. Mizoguchi, Infant drowning prevention system with wireless accelerometer—Evaluation of optimum floating body shape for home-use, in Proc. SENSORS, Lecce, Italy, 2008, pp. 1218–1221.
DOI
[20]

S. Q. Ren, K. M. He, R. Girshick, and J. Sun, Faster R-CNN: Towards real-time object detection with region proposal networks, IEEE Trans. Pattern Anal. Mach. Intell., vol. 39, no. 6, pp. 1137–1149, 2017.

[21]
B. Liu, W. C. Zhao, and Q. Q. Sun, Study of object detection based on Faster R-CNN, in Proc. Chinese Automation Congress (CAC), Ji’nan, China, 2017, pp. 6233–6236.
DOI
[22]

S. Luo, J. Yu, Y. J. Xi, and X. Liao, Aircraft target detection in remote sensing images based on improved YOLOv5, IEEE Access, vol. 10, pp. 5184–5192, 2022.

[23]
Ultralytics, https://github.com/ultralytics/yolov5/tree/v5.0, 2022.
[24]
B. Li, M. M. Fu, and Q. Li, Runway crack detection based on YOLOV5, in Proc. IEEE 3rd International Conference on Civil Aviation Safety and Information Technology (ICCASIT), Changsha, China, 2021, pp. 1252–1255.
DOI
[25]
S. Bock and M. Weiß, A proof of local convergence for the Adam optimizer, in Proc. Int. Joint Conf. Neural Networks (IJCNN), Budapest, Hungary, 2019, pp. 1–8.
DOI
Publication history
Copyright
Acknowledgements
Rights and permissions

Publication history

Received: 27 September 2022
Revised: 15 January 2023
Accepted: 19 January 2023
Published: 30 March 2023
Issue date: March 2023

Copyright

© The author(s) 2023.

Acknowledgements

Acknowledgment

This work was supported by the CAAI-Huawei MindSpore Open Fund and the General Program of Natural Science Foundation of Fujian Province, China (No. 2020J01473).

Rights and permissions

The articles published in this open access journal are distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/).

Return