SHRI G. S.

INSTITUTE OF TECHNOLOGY AND SCIENCE

Department of Biomedical Engineering

MINOR PROJECT REPORT

“ Human Body posture monitoring system”

Submitted to: Submitted by:

Mrs. Avni Jain Mr.Samit Jain
Mr. Sunny Ganavdiya Miss Umang Shrivastava
Mr. Pranshu Paliwal
 ACKNOWLEDGEMENT

In the accomplishment of this project successfully, many people have bestowed

upon us their blessings and heart pledged support, this time we are utilizing to
thank all the people who have been concerned with this project.
Primarily we would thank god for being able to complete this project with success.
Then I would like to thank our Minor project mentor Mr. Vinay Manurkar, our
project guide Mr. Sunny Ganavdiya and our co-guide Mrs. Avni Jain whose
valuable guidance has been the ones that helped us to patch this project and make it
full proof success. Their suggestions and instructions have served as the major
contributor towards the completion of the project.
Then we would like to thank our parents and friends who have helped us with their
valuable suggestions and guidance has been very helpful in various phases of the
completion of the project.
 TABLE OF CONTENT

S.No. Contents Page No.

1. Abstract 4
2. Introduction 5-7
3. Related Works 8-12
4. Materials Required 13-14
5. Proposed System 15-16
6. Conclusion 17
7. Future Enhancement 17
8. References 18
 Abstract

In order to detect and correct the posture we developed a wearable garment

integrated device to sense the posture of the user and analyze the posture within the
device to alert the user remotely. Further the posture data is sent to the server to
analyze the data in detail using a flex sensor.
The posture is analyzed and if bad posture is detected then the buzzer sensor
present in the device alerts the user to correct the posture.The data are stored in the
database along with time, date and device id.The month wise statistical graph is
developed from the data in the database. Daily analyzes of the posture data is done
to provide the average posture of the user and worst posture of the user.
 HUMAN BODY POSTURE MONITORING SYSTEM

INTRODUCTION:
Posture detection and alert system is used to monitor the posture of the users and
alert them to correct the bad posture.Posture is the position in which we hold our
bodies while standing, sitting, or lying down.Posture is not a position, but a
dynamic pattern of reflexes, habits, and adaptive responses to anything that resists
you being more or less upright and functional.
Poor posture soon results in Sciatica problem.Sciatica is a term that describes
symptoms of pain, numbness, and weakness that radiate along the sciatic nerve
from the lower back to the buttocks and leg as shown in(Appendix
1).(Forward-flexed Posture Detection for the Disease Symptom by Wen-Shao Wu,
Wen-Yen Lin and Ming-Yih Lee October 2014).Most commonly, sciatica is caused
by a disc problem, such as a herniated disc that is pressing against a nerve root.

( Appendix 1 )
Good posture is the correct alignment of body parts supported by the right amount
of muscle tension against gravity as shown in (Appendix 2).Good posture helps us
keep bones and joints in correct alignment so that our muscles are used correctly
and reduces the stress on the ligaments holding the spinal joints together,
minimizing the likelihood of injury.

(Appendix 2)

The project aims at providing a wearable monitoring system to monitor the user
posture continuously using the flex sensor and sends the gathered data to the server
by means of the arduino connected with it. By analyzing the data remotely on the
system and providing the alert to the user by means of the buzzer and also
analyzing the data collected in the server and generating detailed reports on the
user posture.
A flex sensor or bend sensor is a sensor that measures the amount of deflection or
bending. Usually, the sensor is stuck to the surface, and resistance of the sensor
element is varied by bending the surface.
Arduino is an open-source electronics platform based on easy-to-use hardware and
software. Arduino boards are able to read inputs - light on a sensor, a finger on a
button, or a Twitter message - and turn it into an output - activating a motor,
turning on an LED, publishing something online.
Arduino is a tool for controlling electronics. Now what do we mean by this term
electronics- there are two general groups in electronics that are INPUTS and
OUTPUTS.
Inputs are devices which gather information and outputs are devices that do things.
So for inputs you can think of all the sensors like temperature sensor,light sensor,
touch sensor, humidity sensor, infrared sensor, distance sensor etc. We are using a
flex sensor in this project as our input device and these types of sensors are read
with the help of an arduino board.
Outputs will be things like DC motors, servomotor, stepper motors, solenoids,
LCD display, LED indicator lights, speakers and many more. We are using the
buzzer in this project as our output device.
Arduino is used to read inputs and control outputs. When we talk about arduino,
we have to look towards three things which it comprises: Arduino hardware,
Arduino IDE and Arduino code.
First if we see the physical component of an arduino that is the arduino boards.
There are several different types of arduino boards available. One of the most
popular arduino boards is the arduino uno which we have used in our project.
All arduino boards share one thing in common, that is they all possess a
microcontroller in them and a microcontroller is basically a really small computer
and this microcontroller is the one which enables you to read the inputs and control
the different output.
Arduino is more than just a hardware, it’s also comprises of software i.e. Arduino
IDE which stands for integrated development environment used to program the
arduino boards. The third part of this arduino trifecta is the code of the arduino.
The code which is written inside the arduino IDE, ultimately gets loaded onto the
microcontroller and this code is basically C and C++ programming language.
RELATED WORKS:

Assessment of human activity and posture with sensors provides insightful

information about the functional ability: classification of human activities in
rehabilitation and elderly surveillance contexts has been already proposed in the
literature. This paper is proposed by DavideCurone, Gian Mario Bertolotti, Andrea
Cristiani, EmanueleLindoSecco,and Giovanni Magenes. In the meanwhile, recent
technological advances allow developing miniaturized wearable devices, integrated
within garments,which may extend this assessment to novel tasks, such as
real-time remote surveillance of workers and emergency operators intervening in
harsh environments. We present an algorithm for human posture and activity-level
detection, based on the real-time processing of the signals produced by one
wearable triaxial accelerometer. The algorithm is independent of the sensor
orientation with respect to the body. Furthermore, it associates to its outputs a
“reliability” value, representing the classification quality, in order to launch reliable
alarms only when effective dangerous conditions are detected. The system was
tested on a customized device to estimate the computational resources needed for
real-time functioning. Results exhibit an overall 96.2% accuracy when classifying
both static and dynamic activities. This paper presents a human activity classifier,
based on the real time analysis of the signals detected with a triaxial accelerometer
fixed to the trunk. Low-level routines, suitable for implementation on a low power
microcontroller, process the raw accelerometer signals in order to extract simple
features, which are directly related to the posture and activity intensity.

Postural change is an important factor in the care and rehabilitation of elderly

patients. Changes in gait can signal the onset of neurological diseases such as
Parkinson’s,whilst changes in how routine tasks are performed – such as getting
out of a chair – can indicate an increase in frailty and a loss of muscle strength
which can increase the chance of a fall . By characterizing motion within the home
it becomes possible to provide a more accurate picture ofelderly patients’ health
and wellbeing. Unfortunately, there are various difficulties associated with
monitoring specific activities, particularly those related to transitions. Firstly, these
motions represent only a small percentage of daily activity. Secondly, they can
occur anywhere within the home and the subject may be oriented in any direction.
Finally, a high level of subject detail and context is required. In this paper, we
present a motion characterization framework based on the fusion of both wearable
and ambient sensors. We use a probabilistic method for detailed posture extraction
which utilizes only four cameras and provides a detailed, but privacy respectful,
rotationally invariant signature of posture. Data from a wearable accelerometer is
fused with optical flow to provide both local and global motion detection which is
robust to the relative positioning of camera and subject. We further discuss our
method for the fusion of gyroscope and video data which provides both the
advantages of detailed local posture features along with a global estimation of
posture through an ear mounted gyroscope. The results derived demonstrate that
complex motions can be detected and distinguished accurately. In this project, we
present a system for the accurate characterization of motion based upon the fusion
of ambient and wearable sensors. A probabilistic, privacy respectful method for the
extraction of detailed 3D posture information is proposed.

With an aging society coming, there are more and more elderly people having the
aging related chronic diseases, such as Parkinson’s, and Alzheimer’s diseases. In
this study, a timeless linear transformation method is proposed to obtain tilting
angles from single axis accelerometer data. By wearing the previously designed
posture monitor vest and with the proposed algorithm implemented, this wearable
system can detect the forward-flexed posture which is frequently seen in the early
symptom of Parkinson’s disease, the festination. Detection of this posture is the
necessary function for the destination detection system which can work as a
quantitative tool for early detection of Parkinson’s disease. In this work, a wearable
forward-flexed posture detection is developed. The upper body lean forward angles
are measured to detect the posture and also to indicate the severity of the posture
by wearing the posture monitoring vest designed previously. Since the upper body
lean forward angles are only within a certain tilting range, a time-less linear tilting
angle transformation algorithm is proposed for fast angle linear transformation
which takes only a small amount of time.

A light-weight online classification method to detect smartphone user’s postural

actions, such as sitting, standing, walking, and running. These actions are named as
‘user states’ since they are inferred after the analysis of data acquired from the
smartphones equipped with sensors. To differentiate one user state from another,
many studies can be found in the literature.

However, this study differs from all others by offering a computational lightweight
and online classification method without knowing any priori information.
Moreover, the proposed method not only provides a standalone solution in
differentiation of user states, but also assists other widely used offline supervised
classification methods by automatically generating training data classes and/or
input system matrices. Furthermore, we improve these existing methods for the
purpose of online processing by reducing the required computational burden.
Extensive experimental results show that the proposed method makes a solid
differentiation in user states even when the sensor is being operated under slower
sampling frequencies.

Changes in the society workforce in the last decades has forced the adult
population to spend long periods of time in a sitting position in the workplace that
coupled with a sedentary lifestyle at home is associated with health problems, such
as back and neck injuries . While a person is seated, most of their body weight is
transferred to the ischial tuberosities, to the thigh and the gluteal muscles. The rest
of the body weight is distributed to the ground through the feet and to the backrest
and armrest of the chair when they are available. The adoption of a lumbar flexion
position for long periods of time, can lead to a decrease of the lumbar lordosis,
causing anatomical changes to the spine and degenerate the intervertebral discs and
joints, disorders that have been linked to back and neck pain. In order to build an
intelligent chair capable of posture detection and correction we developed a
prototype that measures a pressure map of the chair’s seat pad and backrest and
classifies the user's posture. The posture classification was done using neural
networks that were trained for 5 standardized postures achieving an overall
classification of around 98%. Those neural networks were exported to a mobile
application in order to do real-time classification of those postures. Using the same
mobile application we devised two correction algorithms that were implemented in
order to create an intelligent chair capable of posture detection and correction. The
posture correction is forced through the change of the conformation of the chair’s
seat and backrest by changing the pressure of eight pneumatic bladders.
Measurement of human posture and movement is an essential area of research in
the bioengineering and rehabilitation fields. It is motivated by different goals in
clinical application, such as in comparing normal movements with pathological
movements, planning and evaluating treatment protocols, and evaluating design of
orthosis and prosthesis. Human postures and movements have been measured by
using different image-based methods including photogrammetry (Weissman 1968;
Bullock and Harley 1972; Thometz et al. 2000;Liu et al. 2001), optoelectric
analysis (Pearcy et al. 1987; Dawson et al. 1993; Gracovetsky et al. 1995), and
video analysis (Robinson et al. 1993; Masso and Gorton 2000; Nault et al.2002;
Engsberg et al. 2003). The characteristics of different image-based methods were
summarized by Hsiao and Keyserling (1990). Photogrammetric systems have been
used to record two- or three dimensional images of posture. This type of system
uses either light reflective markers or light-emitting diodes affixed to the human
body, and captures data with cameras and films for measuring the orientation of
body segments through data reduction processing. Optoelectric analysis applies the
same principles as photogrammetric system to measure the position of joints and
body segments. The optoelectronic sensing unit is used for collecting the data
instead of films. Video systems also use the similar basic principles as both
Photogrammetric and optoelectronic systems but capture data with optoelectronic
units or cameras of higher sampling rate. These systems can be used to capture and
record three-dimensional body movements. The availability of these image-based
methods has helped to achieve the goals of monitoring and analyzing human
posture and movement. However, inherent limitations of these methods, which are
complicated to set up, time-consuming to operate, and limited to the laboratory
environment (Hsiao and Keyserling 1990), so the chance of using these methods in
the clinical applications are restricted. In recent years, low powered and
miniaturized electronic sensors, which are for use in robotic, industrial, aerospace
and biomedical applications, have been developed by using advanced electronic
circuit technology. The use of these electronic sensors has been considered as
alternative methods for human posture and movement analysis in clinical
applications. The purpose of this article is to review the possible clinical
applications of different types of electronic sensors and systems, and their
problems and limitations which are faced in the human posture and movement
measurements. Such information would help researchers and clinicians in
developing and selecting the most appropriate measurement techniques of using
the electronic sensors for clinical applications of human posture and movement
analysis.

Karaçizmeli et al. (2014) used a flex sensor placed on a glove for a

mechatronic-based robotic hand control. The angular values of the fingers are
sensed and processed by a microcontroller, and they used this data for control of
servo motors and robotic hand controls. In their work, they saw that a robotic hand
could imitate the movement of a human hand put on a glove. This proves that the
robotic hand can be used in contactless and remote environments that can be
harmful to humans in a variety of environments. Korkmaz et al. (2016) designed a
robot arm that mimics the movements of the fingers according to received signals.
Using a 3D printer, a flexible hand glove model was developed by them, a flexible
glove placed on an ordinary glove, this developed system turns motions into
electrical signals and this developed system gives a chance for different servo
motor systems to work together. Tripathi and Pandey (2011) worked on impact of
drudgery reducing technology on ergonomics of rural women engaged in milking
of animals. They refer the correct posture degree affect the heart rate of rural
women
engaged in milking of animals. OWAS, REBA, etc. posture assessment studies are
based on observation measures. In this study, an electronic measurement system
was used for this aim. Also in this study, considering the developments and price
decreases in sensor technology, it is aimed to provide a safety monitoring system
which gives an audible warning by measuring the posture angle which can be
formed on some threshold value, which can be selected during agricultural works.
The data of this system are compared with the image processing technique
measurements. The relationship between the results is explained by the
regression values and the regression equations, and it is discussed in the result
section of the research.
MATERIALS :

For the proposed project components used are:

1. Flex sensor
Specifications -
The Flex Sensor is a unique component that changes resistance when bent.
In this model a flex sensor of 4.5 inch is used.

Flat Resistance 10k ohms

Resistance Tolerance ±30%

Bend Resistance Range 60K to 110K Ohms

Thickness 1mm

Power Rating 0.50 Watts continuous.

1 Watt Peak

Life cycle >1 million

Temperature Range -35°C to +80°C

2. Buzzer
Specifications -

Input Voltage 5V

Resistance 42 ohms
 Resonance Frequency 2048 Hz

Body Size 12 x 8 mm

Pin Pitch 6mm

External Material Plastic

Colour Black

3. Resistor : A resistor of resistance 10k ohms is used in the project model.

4. Jumper wires : Various male to male, male to female and female to female
jumper wires were used made up of sleeve material as high grade copper
wire.

5. Arduino uno board :

Specifications

Microcontroller ATmega328P

Operating Voltage 5V

Digital I/O Pins 14

PWM Digital I/O Pins 6

Analog Input Pins 6

DC Current per I/O Pin 20 mA

DC current for 3.3V Pin 50 mA

Flash Memory 32 KB (ATmega328P) of which 0.5

KB used by bootloader

Clock Speed 16 MHz

 LED_BUILTIN 13

Length 68.6 mm

Width 58.4 mm

Weight 25 g

Input Voltage (limit) 6-20V

PROPOSED SYSTEM:

The Product reminds the user to correct their posture.The circuit diagram for the
system is designed as mentioned below.

A 5v from Arduino is connected with the flex sensor and the ground is connected
with the 10k resistor. This flex sensor and 10k resistor make a voltage divider, a
wire from the middle is connected with the Analog pin A0 of the Arduino. The
buzzer’s S Pin is connected with pin number 5 of the Arduino which is the PWM
pin and the minus pin is connected with the ground.(Posture determination by
using flex sensor and image analysis technique by abdullah beyaz)
The device gets the value from the flex sensor, resistance of the flex sensor
increases as the bending increases.

The arduino code which is programmed in the arduino IDE is shown in (Appendix
3)

13 void setup()
14 {
15 Serial.begin(9600);
16
17 pinMode(flexs, INPUT);
18 pinMode(buzzer, OUTPUT);
19
20
21 }
22
23 void loop()
24 {
25
26 flexdata = analogRead(flexs);
27 Serial.print("flex value;");
28 Serial.print(flexdata);
29 Serial.println("");
30 if( flexdata < 220)
{
analogWrite(buzzer, 150);
}

if( flexdata > 220)

{
analogWrite(buzzer, 0);
}

delay(1000);

(Appendix 3)

User is alerted if the posture of the user is not good. The posture of the user is
calculated from the flex sensor value reading. If the reading is greater than the
threshold value then the user is alerted using the buzzer sensor in the device. The
user is only if the reading is greater than the threshold value for about a minute.
The buzzer is activated for five seconds and goes to idle. The device sends the data
to the server only when the alert system is activated. The normal posture need not
be analyzed.

CONCLUSION:

Posture related problems are a major problem in the modern world. Bad posture
causes many health related problems. Thus our project provides a device to detect
and correct the posture of the user. garment-integrated bend sensing provides an
accurate and reliable measure of seated spinal posture through measurement of
sagittal flexion. In addition, the flex sensor is not too expensive and well suited to
wearable applications, because of its small size, flexibility, and easily-customized
length. In this study, the sensor was evaluated in a wearable configuration designed
to minimize the number of additional influencing variables. Removing these
variables, however, often requires sacrifices in user comfort and social
acceptability of garments.

FUTURE ENHANCEMENT:

Future work will include analysis of the effects of garment ease and fit on the
sensor’s performance. Ultimately, this sensor will be integrated into a more
complete biofeedback posture monitoring system. By integrating multiple sensors
in the device the posture of the user can be measured with more accuracy. By
reducing the size of the sensor and battery user comfort can be increased.
REFERENCES:

1. Clinical applications of sensors for human posture and movement analysis:

A review by WAI YIN WONG, MAN SANG WONG, & KAM HO LO
March 2007.
2. Forward-flexed Posture Detection for the Early Parkinson’s Disease
Symptom by Wen-Shao Wu, Wen-Yen Lin and Ming-Yih Lee October 2014.
3. Posture determination by using flex sensor and image analysis technique by
abdullah beyaz.
4. Hsiao and Keyserling (1990), (Robinson et al. 1993; Masso and Gorton
2000; Nault et al.2002; Engsberg et al. 2003), (Pearcy et al. 1987; Dawson et
al. 1993; Gracovetsky et al. 1995), (Weissman 1968; Bullock and Harley
1972; Thometz et al. 2000;Liu et al. 2001)