Automated drowning system for swimmers: a review

Abstract

Every year, many individuals are drown in the deeps of the beach, and the lifeguards are not well trained enough to handle these situations. Thus arises the requirement for having a system that will consequently detect the drowning individuals and alarm the life guard at such risk. A Beach surveillance system plays an essential role in safeguarding the premises. In this project differential pressure approach is used for detection of drowning incidents in beaches at the earliest possible stage. The people life is saved during drowning incidents in the beach by wearable device. The proposed approach consists of Wearable module, Standstill module and Monitoring system. The prototype system based on health monitoring sensors has an advantage of convenience, lifesaving and real time.

Keywords — Drowning, biomedical sensors, wearable, GPS, ZigBee module, Repeater

I. Introduction

There are many safety, security, detection and tracking devices implemented for children, women’s, environment, car tracking, houses, remote monitoring, etc. Many times people lose their lives from drowning in water; this is not an exception for the great swimmers also. The World Health Organization (WHO) classifies drowning as the 3rd leading cause of unintentional injury worldwide after road traffic injury. Everyone admires playing at beach and is a great stress buster. But in the water, beginners often feel hard to breathe which causes choking actions, loss of balance and results in a drowning accident. Some special circumstances, such as cramps, collide with each other, disease or mental stress and so on may also cause drowning. Many times swimmer gets struck or any tragedy happens which drive them towards drowning in water. Sometimes due to high pressure of water person gets flow away due to which it becomes difficult to track a person.

Thus automatically understanding events happening at a site is the ultimate goal of surveillance system. This paper investigates the challenges faced by automated surveillance systems operating in hostile conditions and demonstrates a developed prototype system that detects life threatening situation of humans within highly dynamic aquatic environments. The developed wearable system is incorporated into a live system with robust performance for different hostile environments faced by an outdoor beach.

II. Literature Survey

One of the previous systems proposed by S. Hemalatha, P. Nandhini, J. Vimala, and V. Ramesh [1] was designed by using a light source of LDR and laser placed on the side of the wall to identify humans. The iron metal plate was placed in the floor of the swimming pool which was lifted automatically using the motor and motor drive. The laser source was allowed to spread over the swimming pool and the LDR to sense the laser light which produced the resistance value and when LDR value was constant, the alarm was activated. The resistance value was changed with respect to the human movement and message was send to the administrator by using the GSM service.

Second system designed by A Kanchana, Kavya G. R, Kavitha C, Soumyashree V, Salila Hegde [2] was especially for children. This automated drowning detection system worked on the principle of differential pressure. The system had two fundamental modules: to begin with, the wristband consisting of pressure sensors on the transmitter side which should be worn by every child entering the swimming pool and second is the receiver module at the swimming pool site. The Pressure at underwater is different and greater than the pressure at the air – water interface. The pressure at a particular depth was set as the threshold. When the current value transcends the threshold limit an alerting signal was sent to the receiver of RF module. On receiving the valid signal, controller sets the buzzer ON, turns ON the motor driver which in turn lifts the acrylic plate of the multi-floored swimming pool and the kid is brought to air-water interface, i. e. the top level of swimming pool.

Poseidon is a video based drowning detection system reported in literature survey by A Kanchana, Kavya G. R, Kavitha C, Soumyashree V, Salila Hegde [2]. This system has three kinds of drowning monitoring system according to the different position of the camera. One is camera mounted underwater, other camera is mounted upon the water while third is a combination of the two. These cameras monitor the swimmer status and posture change. A limitation of this equipment is that if too much swimmers, occlusion problem will appear. The reaction and refraction of light and water wave interference will affect the image quality, and drowning man feature this method detected is not easy to distinguish swimmers and divers obviously. This system needs constant observation, expensive installation costs which are the main disadvantage.

The system proposed by Johan Carlén, Jonas Larsson [3] had wristband which consisted of several system components, which together form a network of sensors, processing units, alarm triggers and wearable communication units and mainly were dependent on RFID-chip standards for supporting the Sen-tag system. The wristband senses pressure and trigger an acoustic signal in the water. For safety reasons, there was always two hydrophones one in visual contact with the wristband and other in the swimming pool. The wristband is equipped with a RFID chip so it can communicate with access points within the facility. A prototype of wearable smart locator band was designed by Isha Goel and Dilip Kumar [4] which can be worn on the wrist of the children to monitor and keep an eye on them. The developed device includes an AVR microcontroller (ATmega8515), global positioning system (GPS), global system for mobile (GSM), and switching unit and the monitoring unit includes Android mobile device in parent’s hand with web based Android application as well as location indicated on a Google Map.

This development was very useful for senior people and individuals suffering from memory diseases . This device, hence, behaved as a communication interface between wearer and caregiver. One of the works involved a wearable device for the safety and protection of women and girls proposed by Anand Jatti, Madhvi Kannan, Alisha RM, Vijayalakshmi P, and Shrestha Sinha [5]. This objective was achieved by the analysis of physiological signals in conjunction with body position. Real-time monitoring of data was achieved by wirelessly sending sensor data to an open source Cloud Platform.

Analysis of the data was done on MATLAB simultaneously. This device was programmed to continuously monitor the subject’s parameters and take action when any dangerous situation presents. It did so by detecting the change in the monitored signals, following which appropriate actions were taken by means of sending notifications/alerts to designated individuals. Acquisition of raw data was followed by activity recognition which was a process of employing a specialized machine learning algorithm. The health status of human body can be indicated by a variety of physiological parameters, which is stated by Yan Liu, Hai Wang, Wei Zhao, Min Zhang, Hongbo Qin and Yong Qian Xie [6].

The human parameters are evaluation and categorized as:

1. body motions, including hand, limb, foot, face, throat etc.,
2. vital signs, including breath/heart rate, wrist pulse, ECG, blood pressure, skin temperature, SpO2, etc.

These parameters have been detected by various wearable sensors The reliance on underwater cameras in one of the systems put forth by How-Lung Eng, Kar-Ann Toh, Alvin H. Kam, Junxian Wang and Wei-Yun Yau [7]. They summarized as,

1. Expensive installation costs
2. Drowning detection being constrained to victims who have sunk to the bottom of the pool.

To circumvent these drawbacks, the proposed system is based on a network of highly mounted overhead cameras. This allows the detection of early drowning behavior from the onset of water crisis situation. Hence, any rescue effort could be initiated much earlier than those in water. One paper by Mohamed Kharrat, Yuki Wakuda, Shinsuke Kobayashi, Noboru Koshizuka, Ken Sakamura [8] described a swimming cap that detects drowning situation as early as possible by analyzing the swimmers physiological states. A person drowning tends to have a vertical body posture while struggling at the same location to ensure that the swimmer is in danger.

The heart rate activity can be analyzed from ECG or PPG signals. For the body posture estimation, the vertical position can be detected using an accelerometer or gyroscope. The information will be processed in the cap and an alarm will be triggered if an abnormal behavior is detected. But one major technical challenge faced is to accurately detect and track swimmers within the noisy outdoor aquatic environments which are not feasible. In paper [9] proposed by Nasrin Salehi, Maryam Keyvanara, Seyed Amirhassan Monadjemmi a method is provided to robust human tracking and semantic event detection within the context of video surveillance system capable of automatically detecting drowning incidents in a swimming pool.

In the current work, an effective background detection that incorporates prior knowledge using HSV color space and contour detection enables swimmers to be reliably detected and tracked despite the significant presence of water ripples. The system has been tested on several instances of simulated water conditions such as water reflection, lightening condition and false alarms. Our algorithm was able to detect all the drowning conditions along with the exact position of the drowning person in the swimming pool and had an average detection delay of 1. 53 seconds, which is relatively low compared to the needed rescue time for a lifeguard operation. Our results show that the proposed method can be used as a reliable multimedia video-based surveillance system

References:

1. S. Hemalatha, P. Nandhini, J. Vimala, V. Ramesh; Automated Drowning Detection and Security in Swimming Pool; IJIRCCE (An ISO 3297: 2007 Certified Organization) Vol. 3, Special Issue 2, March 2015
2. A Kanchana, Kavya G. R, Kavitha C, Soumyashree V, Salila Hegde; Automated Drowning Detection and Security in Swimming Pool; (IRJET) e-ISSN: 2395 -0056 Volume: 04 Issue: 06 | June -2017 www. irjet. net p-ISSN: 2395-0072
3. Johan Carlén, Jonas Larsson, Civilingenjör, Teknisk design 2017; Luleå tekniska universitet, Institutionen för ekonomi, teknik och samhälle.
4. Isha Goel, Dilip Kumar; Design and Implementation of Android Based Wearable Smart Locator Band for People with Autism, Dementia, and Alzheimer; Received 7 August 2014; Revised 1 December 2014; Accepted 9 December 2014; http://dx. doi. org/10. 1155/2015/140762
5. Anand Jatti, MadhviKannan , Alisha RM, Vijayalakshmi P, ShresthaSinha; Design and Development of an IOT based wearable device for the Safety and Security of women and girl children; IEEE International Conference On Recent Trends In Electronics Information Communication Technology, May 20-21, 2016, India
6. Yan Liu, Hai Wang , Wei Zhao, Min Zhang, Hongbo Qin, Yongqiang Xie; Flexible, Stretchable Sensors for Wearable Health Monitoring: Sensing Mechanisms, Materials, Fabrication Strategies and Features; Received: 18 December 2017; Accepted: 16 February 2018; Published: 22 February 2018; Sensors; Review
7. How-Lung Eng, Kar-Ann Toh, Alvin H. Kam, JunxianWang, Wei-Yun Yau; An automatic drowning detection surveillance system for challenging outdoor pool environments; Proceedings of the Ninth IEEE International Conference on Computer Visison (ICCV 2003) 2 -Volume Set; 0-7695-1950-4/03 $17. 00 © 2003 IEEE
8. Mohamed Kharrat, Yuki Wakuda, Shinsuke Kobayashi, Noboru Koshizuka, Ken Sakamura; Near drowning detection system based on swimmer’s physiological information analysis; World Conference on Drowning Prevention 2011; ID: 18/Paper: 370
9. Nasrin Salehi and Maryam Keyvanara Seyed Amirhassan Monadjemmi, “ An Automatic Video-based Drowning Detection System for Swimming Pools Using Active Contours”, MECS I. J. Image, Graphics and Signal Processing, 2016, 8, 1-8
10. World Health Organization (WHO), Drowning Fact Sheet (2018), http://www. who. int/mediacentre/factsheets/fs347/en/