ABSTRACT Every day we are seeing a lot of accidents happening around the world. In 2015 Tamilnadu records the highest percentage of people who met with traffic accident injuries. An effective way for reducing these traffic fatalities is to first develop automatic traffic accident detection system and second is reducing the time when first emergency responders reach that place. Nowadays BMW is manufacturing vehicles with an inbuilt automatic accident detection and notification system. Even though this is a better approach and may work fine but this is very much expensive, complex and may not be available for all the vehicles and the maintenance costs are also very high comparatively.
Nowadays we can detect traffic accidents using smart-phones and that ability has only recently become possible because of the technological advancements such as different sensors deployed in our smart-phones and their processing power. In our proposed work we use smart-phones for accident detection. The whole system relies on the G-Force value (extracted from smart-phone accelerometer sensor) to detect an accident in the offline mode. To elucidate the fall detection technique, we use the waveform sequence of the tri-axial accelerometer (X, Y&Z Axis) on the smart-phone as the system inputs. The acquired signals are then compared with our threshold G-Force value so that it can fetch the GPS location of the smart phone thereby track the victim. The application consists of two phases, the detection phase which is used to detect an accident and the notification phase which is used to send a detailed information such as accident location longitude and latitude coordinates, an emergency message to the emergency responders (emergency contacts that were set earlier) when an accident happens. This will be useful for the fast response as the person who received those help notification can immediately call the nearby hospital and can request for emergency service. The message(SMS) is generated based on phone tower signal and if the phone is switched off in any case then the last updated location will be sent.
Here we use the GSM technology to send the message and the location via SMS. The proposed android application can work well even in low signal strengths and can also run in the background. Keywords: Tri-axial Accelerometer Sensor, G-force value, Emergency contacts, Latitude and Longitude Coordinates. I. INTRODUCTION Nowadays the accident rate has been gradually increasing every day along with the rapid development of a transportation system. The automobile has a great importance in our daily life. We use it to go to our workplace, keep in touch with our friends and family, and deliver our goods.
The rate of automobiles has been increasing by 6% every year. But it can also be a disaster to us and even can kill us through accidents. In 2017 Traffic accident record shows that 1250 road accidents happen every day in India 400 people die every day only due to road crashes. In 2016, TamilNadu registered a total of 69,059 accidents, nearly 14 percent of all road accidents in India. Speed is one of the most important and basic risk factors in driving. It does not only affect the severity of a crash but also increases the risk of being involved in a crash.
So our proposed work can help such victims and reduce the fatality rate. We basically develop an android application for accident detection and notification. Nowadays everyone is using smart-phones because of its low cost and high features and also most of the people frequently change their smart-phones when compared with vehicles. The advantage of using smart-phones for detecting accidents is that the technological innovations have been improving day by day with new software and hardware like sensors, communication interfaces etc which helps in performing various tasks and in our proposed work it helps in detection and notification during accidents. In this paper, we have implemented an android application that runs from Android 4.0 till the latest Android 7.1.1(Nougat).
The accident detection can be done by using various sensors deployed in our smart-phone such as triaxial accelerometer and also by using human fall detection technique. This triaxial accelerometer calculates the G-force value in X, Y & Z axis and if the G-force value greater than the threshold value then the emergency message will be sent to our emergency contacts. The coordinates at that particular axis can be fetched using the tower signal from the base-stations.
The GSM technology can be employed so that there is no need for any internet connectivity. The already existing work related to this concept is to calculate the G-force value using the 3G/4G internet connections. And this model works only when connected to the internet and another related existing work is in which some wearable sensors which we found as durable as it is not comfortable to everyone to carry the sensors to each and every place and the cost of this method is also much higher due to the usage of many sensors. Our Proposed work has two phases the detection and notification phase the detection phase used to detect the abnormal threshold value and capture the place Co-ordinates that is Longitude and latitude points and the notification system use the service providers cellular network signal and sends the GSM precisely it will send SMS to our emergency contacts without the use of any GPRS,3G etc. In our proposed work we have developed an android application in which we develop a sign up activity and login activity and we create a local database and link the login contacts to the database and if the application is started after the login activity the G-force value is calculated and if any abnormal ratios are identified greater than the fixed threshold value then the message and the coordinates will be sent through SMS and if in any condition the smart phone breaks due to severe accidents the last travelled location of the mobile network captured and sent to the emergency contacts. In the previous works, the android application works only when the application is ON and it not running in the background. But we implemented the application even which works in the background. Rest of our paper consists of related work in section II, Implementation in section III, Conclusion in section IV.
II. RELATED WORKS The innovation in the technology helps to reduce the rate of accidents by using Android platforms and smart-phones2. In this, they proposed that 3G internet connectivity along with the mobile GPS tracker in the smart-phones are used to locate the latitude and longitude coordinates, accident data recording and takes photos for situational awareness.
These coordinates are sent to the pre-registered contacts in abnormal conditions. An issue arises here that the internet access will not be available to every user.Nowadays falls became a serious public issue, it might be an accident or a sudden fall due to unconsciousness those falls are detected with wearable sensors fitted to subjects body1. In this paper, they the fall among the elderly persons is detected. They worked using triaxial devices like accelerometer sensor, magnetometer, gyroscope and also employed 14 volunteers to perform some set of actions. Those actions helped them in identifying the cause and position of fall.
Some wearable sensors are very difficult in carrying all the time.Smart-phones plays a crucial role in our day to day lives. While travelling if some people met with an accident or lost somewhere else they can be helped by their caretakers by sending location coordinates an emergency message via email, MMS or tweets about the accident scenario through an android application3. They proposed that accidents or human falls are detected using integrated sensors. As not everyone checks their email and twitter all the time and also it does not run in the background there will be some issues in this work.The existing works on Android applications called CALL AMBULANCE Where smart-phones are used to detect and notify the accidents.It has an HELP button in the user interface.
The patients who are prone to the accidents may not be able to click that HELP button.The population of seniors has been growing in our society.If they met with some unusual events they have to check with the previously worked falling positions4.
If the fall matches then the cause of elderly persons in the video surveillance can be easily determined. The video surveillance will be a difficult task as we have to monitor the elderly person every time. It is not accurate and may even give the false results.
After the thorough exploration of all these works, we have proposed a system with its implementation in a smartphone-based application available in offline mode. It is known for its ease of portability, make the users quite comfortable to use and effective in detecting the accidents. we created an application that runs in the background and makes it fetch the GPS coordinates of the last updated location in case of any phone damage. This application can be used for fast emergency rescue and accurate results. III.
IMPLEMENTATION OF OFFLINE SMART-PHONE BASED HUMAN-FALL DETECTION SYSTEM Our proposed work was made to detect accidents using smart-phones, here if an accident occurs to any person then immediately the inbuilt smart-phone sensors such as accelerometer sensor used to detect the falling ratio along with all the axis. For suppose if an accident occurs the G-Force value will be calculated along the three axis(X,Y &Z) let it be 1.5g,4g,6g(Say) then the G-Force(6g) value at a particular axis which is greater than the threshold value(5g) will be considered as an abnormal situation and immediately our application fetches and sends the GPS location along with the emergency message to the pre-registered emergency contacts. In case of a sudden jerk or brake, the G-force value will be very minimum so that it will be taken as false accident scenario.
We can set the threshold limit after which the messages will be sent. Fig 1:Architecture diagram of Offline based smart-phone human-fall detection system Fig.1 describes the entire process of our proposed system. Here we have four activities namely splash, user sign in and sign up, database creation and run activity. Here we use XML for user interface design. All the above-mentioned activities have java file and a design page. Initially, we have splash activity which is used for beautification purpose, here we can set the thread for 5 seconds(Say).
The thread will be destroyed as soon as the timer ends. The intent is used to move from one activity to another. In sign-in and sign-up activity we have to enter username and password, then immediately it will have a check in a database(SQL Lite database created using database adapter).
If the user is not existing it will immediately prompt an error message. For a new user, we have to register with a username and password. In database, it is easy to add, delete, update the data as it is stored in the form of tables. Here we will give our emergency contacts and emergency message then the run activity will be started now the detection phase will be done using the inputs from the triaxial accelerometer, if the G-Force value is greater than the threshold value then the SMS will be sent. Fig 2:Splash Activity Fig3: Sign-in ,Sign-up and Database Activities Fig 4:Run Activity Our application is implemented in such a way that a sender(user 1)sends notifications to the receiver side(n users).
We are implementing such that even at a 2g signal the SMS will be sent. In the below fig.5, we have a sender(user 1)who log in to our application and register with some emergency contacts and message in the start session, Here the GPS location fetching is done in the background, and in the foreground we will check for an abnormal motion variations(inputs from accelerometer sensor). If the situation is abnormal the current location and message will be fetched and sent to the receiver. Here we have used GSM to search the nearby base station to send the SMS.
Fig 5: Process of sending notifications about an accident via SMS(Latitude&Longitude coordinates and Emergency text) to the receiver. IV. CONCLUSION Our proposed work represents an Android Application which is offline smart-phone based human-fall detection system. The inputs in our implementation are taken from the integrated smart-phone sensors like the accelerometer. Our main work includes that the inputs are compared with the prefixed threshold value and the maximum value along a particular axis will be used to capture the location coordinates. We are working to implement our application to run even in background such as Google maps. We implemented the freeze option which will be helpful to run the application to run smoothly even when calls interrupts. This application can also be made to be more user friendly by making the user interface in many languages so that any kind of people can understand the functionality of the application.
This helps to save lives of many people by reducing the time between the accident occurance and the emergency responders. ACKNOWLEDGEMENT It gives us immense pleasure to express our profound gratitude to our Department of Computer Science and Engineering, SRM Institute of science and technology, Kattankulathur, for encouraging and giving us support and good environment to gain valuable experiences. REFERENCES 1 Ahmet Turan Ozdemir and Billur Barshan, “Detecting falls with wearable sensors using machine learnin techniques”;Sensors 0691-10708;dio10.
3390/s140610691,2014. 2 Jules White, Chris Thompson, Hamilton Turner, Brian Dougherty, Douglas C. Schmidt. “WreckWatch: Automatic Traffic Accident Detection and Notification with Smartphones”; Springer Science+Business Media, LLC 201,22 March 2011. 3 Gokulnath, U.
V.Anbazhagu, Regan Paulraj, Manigandan and M.Latha ” Smartphone based accident detection system”; International Journal of Pure and Applied Mathematic Volume 114 No.
12: pp:201-211, 2017. 4 Caroline Rougier, Jean Meunier, Alain St-Arnaud, and Jacqueline Rousseau,”Robust Video Surveillance for Fall Detection Based on Human Shape Deformation” IEEE transactions on circuits and systems for video technology, vol. 21, no. 5, May 2011. 5 A.Al-Mazloun, E.Omer, M.
F.A.Abdullah “GPS and SMS based child tracking system using smartphone”; World Academy of science, Engineering and technology International journal for electrical, energetic, computer, Electronic and communication engineering vol:7, no:2, 2013. 6 Rashidi P. Mihailidis.A, “A survey on ambient-assisted living tools for older adults”, IEEE journal of biomedical and health informatics 17(3), pp:579-590, 2013.
7 Zhang, M.; Sawchuk, A.A. “Human daily activity recognition with sparse representation using wearable sensors”; IEEE J. Biomed. Health Inform. , 17, 553–560,2013. 8 Bagalà.
F , Becker.C, Cappello.A, Chiari.L, Aminian.
K, Hausdorff J.M., Zijlstr, W, Klen, J. “Evaluation of accelerometer-based fall detection algorithms on real-world falls.
, 7, e37062, PLoS One 2012. 9 E. Auvinet, L. Reveret, A. St-Arnaud, J. Rousseau, and J. Meunier, “Fall detection using multiple cameras,” in Proc.
30th Annu. Int. Conf. IEEE Eng. Med. Biol.
Soc., pp. 2554–2557, Aug. 2008.