Imperial Journal of Interdisciplinary Research (IJIR)

Vol-2, Issue-11, 2016

ISSN: 2454-1362, http://www.onlinejournal.in

Accelerometer Based Hand Gesture

Controlled Wheelchair
Paridhi Gupta[1], Prachi Sharma[2], Rahul Kr. Singh[3] & Sneha Yadav[4].
Under guidance of Ms. Alpna Singh (Assistant Professor, ECE)
Department of Electronics and Communication Engineering,
Moradabad Institute of Technology, Moradabad (U.P.)

Abstract: In today’s world, physically handicapped This paper proposes a system that can assist the
person & elder peoples are depending on others disabled people to control the motion of their
peoples. But today’s world becomes fast, everyone wheelchair by the hand movements wirelessly. The
is very busy & there are few peoples to take care of system proposed can be mounted on primary
these peoples properly. They find the automated functioning body part to control the wheelchair
wheelchairs for an easy transportation for these movement i.e. hand.
physical disable persons. The proposed work is to
design & develop a hand gesture based wheelchair
using Gesture Control System. Wheelchairs are
used by which person who cannot walk due to
physiological, injury or any disability. Recent
development promises a wide scope in developing
hand gesture based wheel chair or smart
wheelchairs. The present article presents an
accelerometer sensor using gesture based
wheelchair which controls the direction of
wheelchair using hand movements. This paper
presents a model for hand Gesture controlled user
interface and identifies trends in current
technology, application and usability. We present
very useful an integrated approach to real time
detection, hand gesture based data glove technique
is very used which controls the wheelchair using Fig. 1: Pie chart showing cause of paralysis.
hand movements by using the accelerometer
sensor. This paper proposed a low voltage supply, Modified features in the wheelchairs are as follows:
low-cost and small 3-axis wireless system to
control the wheelchair using AVR microcontroller. • Depth sensing feature: This proposed work is
to design an embedded system for
Keywords: microcontroller, accelerometer sensor, implementing an efficient wheelchair that
wheel chair control, hand gesture recognition, RF would move itself when this mode is on. It has
module. some addition feature of DEPTH SENSING,
in which the wheelchair would stop when is
1 Introduction: nearby any pit or staircases. It uses IR sensors
and also LM339 comparator in the sensing of
According to a study conducted by Christopher &
depth obstacles.
Dana Reeve Foundation, nearly every 1 person in
50 is suffering from paralysis due to damaging of • Alcohol detection to prevent accidents: This
nervous system. This figure approximates to 6 proposed work is to design an embedded
million people worldwide and has increased by 33 system for implementing an efficient alcohol
percent from previous estimation [14]. The causes detection system that will be useful to avoid
of Paralysis are mainly due to spinal cord injury accident. In this module we design a system to
stated in the figure. We are using accelerometer in detect the alcohol in breathing and make
our system, accelerometers are also used earlier to wheelchair to stop. If it is conformed the
monitor patients physical activities and their message is send to the nearby control room
posture. [13] using GSM.
• Obstacle avoidance feature: It is one of the
most important aspects of mobile wheelchair.

Imperial Journal of Interdisciplinary Research (IJIR) Page 944

Imperial Journal of Interdisciplinary Research (IJIR)
Vol-2, Issue-11, 2016
ISSN: 2454-1362, http://www.onlinejournal.in
Without it wheelchair movement would be
very restrictive and fragile. Obstacle avoiding
is task which is used for detecting the object
placed in the path of our wheelchair thus it
protects the wheelchair from any damage. An
obstacle avoiding is an intelligent device,
which can automatically sense and overcome
obstacles on its path using Analog IR sensor.
2 Literature survey:
Several attempts have been made in the field of
wheelchair technology to provide the proper
assistant to the needy people. However even these
significant advances haven’t been able to help these
people. The current electric powered wheelchairs
(EPWs) are mostly joystick driven, and cannot
fully meet the need of the disabled and elderly
people whose autonomies are seriously affected by
decline in their motor function and cognitive
performance. Up to now, various hands-free HMIs
have been developed for the disabled and elderly
people to control EPWs by using shoulder, head
and tongue motion, as well as eye tracking.
A. Head Mounted Assistive Technology The head
mounted assistive technology is developed for such
a person who is suffering from quadriplegia but has
retained certain kind of neck or shoulder
movements. This technology is mainly used to
make devices that imitate computer mouse using
head movement. In one of the approach, an infrared
beam emitted or reflected from a transmitter or
reflector attached to user’s cap, glasses or
headband is tracked, for controlling the movement
of cursor. In one of the system tilt sensors are used
to move the cursor in vertical and horizontal
direction on computer screen. To detect the tilt
from the gravity vector, the tilt sensor generally
makes use of inertia [2]. As the user move the head
in some direction the angle between a sensing axis
and a reference vector that is earth’s magnetic field
is detected by sensor and depending upon the angle
cursor movement will take place. One more system
known as ―Camera Mouse has been developed to
control the movement of cursor. This system
consist of a video camera in front of the user,
which continuously track the head or nose
movement and proportionally moves the mouse
pointer on the computer. Jia et al. [3] developed a
visual based HMI for controlling a wheelchair by
head gestures which were recognized by detecting
the position of the nose on user’s face. The main
drawback of these systems is that it constantly
requires neck or shoulder movement which is tiring
and uncomfortable for the user. The head of the
user should always be in the range of sensor
otherwise the user cannot be able to control the
Imperial Journal of Interdisciplinary Research (IJIR)
movement of cursor. Also the design limits the
allowable commands to be generated.
B. Eye Tracking Assistance In eye tracking
assistive technology instead of using a camera for
tracking the eye movement several light sensors
can be used [1]. The sensors can be fitted in eye
glass type apparatus that can be easily wearable by
the user. Also, the light sensors are cheap in cost
hence, reducing the whole system cost.
Furthermore since the used light sensors are small
in numbers, the generated vectors dimension is
small and easy to calculation and fast as well
instead of size of typical video frame captured by
the camera decreasing the computational
complexity of the system. In eye tracking assistive
technology an Eye touch system is used. Eye touch
system consists of components such as infrared
light sensitive apparatus such as infrared light
sensitive apparatus, a data acquisition unit,
computer software and the power supply. Gajwani
and Chhabria[4] used eye tracking and eye blinking
obtained by a camera mounted on a cap to control a
wheelchair. However, the performances of these
HMIs are likely affected by environmental noises
such as illumination, brightness, and the camera
position. Additionally, eye tracking may force and
affect the vision of the user, causing tiredness and
dizziness. Ericka Janet Rechy-Ramirez et al.
developed an intelligent wheel chair which uses An
EEG device, namely Emotiv EPOC to be installed
on used head that make it bulky and complex
[5].This paper is inspired from a research paper
titled “Head Gesture Recognition for Hands-free
Control of an Intelligent Wheelchair” by Jia et al.
[3]. This paper approach was to control a wheel
chair using visual recognition of head gesture.
C. Neural Interface Assistive Technology Recent
advancement in neuro-technology has helped the
users who cannot benefit from mechanical
movement of any body organs by developing the
series of devices known as Brain Computer
Interfaces (BCI) or Neural Interface System (NSI).
This system control the external devices by
detecting user’s intention by utilizing electric
signal originated from brain wave. Recently, a new
EEG sensor, Emotive EPOC, has been available on
the market to provide potential applications on
hands-free HMIs [5]. It has three suites: “cognitive
suite” to detect thoughts, “expressive suite” to
detect facial expressions and “affective suite” to
detect emotions, as well as a gyroscope to detect
head movements. It was used to recognize four
trained muscular events to steer a tractor: eyes
looking to the right and jaw opened, eyes looking
to the right and jaw closed, eyes looking to the left
and jaw opened, and yes looking to the left and jaw
closed. In such systems for making good electrode
Page 945
Imperial Journal of Interdisciplinary Research (IJIR)
Vol-2, Issue-11, 2016
ISSN: 2454-1362, http://www.onlinejournal.in
contact with the cortical region considerable
amount of time is required because if good contact
is not possible the electrode is removed and
cleaned each and every time until we create good
electrode contact. Hence it set up time is more.
Also this system is prone to high error rate as the
signal generated by the EEG has very small
amplitude. The system can give unwanted
command to the device due to interference of
signals resulted from other activities such as talking
and muscle contractions. To overcome such
problem heavy signal processing and complex
computational algorithm can be used, but it will
increase the cost and delay.
D. Voice Controlled Assistive Technology Voice
controlled mechanism can also be used to operate
power wheelchair by the individual who can
produce consistent and distinguishable voice. This
technology makes use of speech recognition system
for taking voice of the user as the input signal.
Before this speech recognition system is used for
actual control of the wheelchair it has to be trained.
The set of command spoken by the user will be
saved by this system. During operation the user
speaks a command into the microphone; the speech
recognition system will compare the spoken
command with the saved command and will
transmit the computer code associated with it. In
this way we can operate the wheelchair. The
advantage of this technology is that, users don’t
have to physically operate the wheelchair and it is
also easy to learn. Out of all these assistive
technologies which were developed, very few
assistive technologies has been proved successful
in outer environment rather than in research
laboratories. There are various technical and
psychophysical factors which affect the acceptance
rate of an assistive technology which are as
follows:
• it should be easy and convenient to
operate
• device should require less time to learn
• it should be cosmetically suitable
3 Proposed Work:
Instead of using all these existing and above
discussed technology we can use a MEMS sensor
that would detect the tilt according to the hand
movement of user and provides RF signal to the
microcontroller that controls the direction
movements of motors along with
• An ultrasonic sensor used for obstacle
detection.
• A depth sensing feature for blind for their
assistance.
Imperial Journal of Interdisciplinary Research (IJIR)
• An alcohol detection device to prevent
accident while using wheelchair on roads.
This would make wheelchair an assistive tool in a
whole manner.
4 Methodologies
As referred earlier this paper comprises of various
modules to work upon with to form a full purpose
and efficiently designed assistive wheelchair for
physically handicaps. The project is been divided
into sub modules which are been explained below:
Module A: Accelerometer based hand Gesture
controlled wheelchair
In this module we will design accelerometer based
hand gesture controlled wheelchair with AVR
ATmega89S52 microcontroller. Here, we will use
our hand gestures as input signals to drive the
wheelchair in different direction and we will
display the direction of movement of the robot in a
16X2 alphanumeric LCD.
Fig. 2: View of hand movement controlling.
In our case, hand gestures mean tilt of hand in
different direction. To detect different hand
gestures, we will use the ADX335 accelerometer
sensor. The sensor will be attached to our hand
with some material or through a hand glove. The
three output signals of accelerometer sensor are
analog in nature and it cannot be processed directly
by ATmega89S52 microcontroller. For this, we
will use the ADC of the ATmega89S52
microcontroller to convert the analog signals to
digital values. After converting the analog signals
of accelerometer sensor to digital values, the
ATmega89S52 microcontroller will process the
digital values to find different gestures of the hand.
Once the hand gesture is known, the
ATmega89S52 microcontroller will send the
required signal to the DC motor driver (L293D) of
the wheelchair to drive the wheelchair in the
desired direction. Also, the microcontroller will
display the direction of movement of the
wheelchair in a 16X2 alphanumeric LCD. The
Page 946
Imperial Journal of Interdisciplinary Research (IJIR)
Vol-2, Issue-11, 2016
ISSN: 2454-1362, http://www.onlinejournal.in
control hand gestures for the wheelchair are
forward tilt, backward tilt, left tilt, right tilt and no
tilt to drive the robot in forward, backward, left,
right direction and to stop it respectively.
Fig. 3: Block diagram of Transmitter.
Fig. 4: Block diagram of Receiver.
Module B: Depth sensing feature.
In this project, the transmitter section includes an
IR sensor, which transmits continuous IR rays to be
received by an IR receiver module. An IR output
terminal of the receiver varies depending upon its
receiving of IR rays. Since this variation cannot be
analyzed as such, therefore this output can be fed to
a comparator circuit. Here an operational amplifier
(op-amp) of LM 339 is used as comparator circuit.
When the IR receiver does not receive a signal, the
potential at the inverting input goes higher than that
non-inverting input of the comparator IC (LM339).
Thus the output of the comparator goes low, but the
wheelchair does not moves. When the IR receiver
module receives signal to the potential at the
Imperial Journal of Interdisciplinary Research (IJIR)
inverting input goes low. Thus the output of the
comparator (LM 339) goes high and the wheelchair
starts moving.
Module C: Alcohol detection device.
Drunken drive is the major problem in every part of
the nation. Many accidents happen due to the
carelessness on the part of the driver. To avoid
accident we designing a system which will assist
the traffic police officers to determine whether
he/she is fit to drive or not. This system is basically
an embedded system which can perform some
specific function using microcontroller AVR-
ATmega89S52. The alcohol sensor detect the
alcohol concentration will give the analog resistive
output to the microcontroller. Microcontroller
controls the dc motor using L293D driver circuit.
Thus vehicle will be stopped on detecting alcohol
concentration and related information will be pass
to the nearby control room using GSM.
The power supply is given to the VCC pin of the
microcontroller AVR- ATmega89S52, the alcohol
sensor MQ3 will detect the alcohol concentration &
provides an analog resistive output which is given
to the inbuilt ADC of microcontroller. The inbuilt
ADC of the microcontroller will convert this
analog resistive output into digital output that will
in the form of LCD message display that
“ALCOHOL DETECTED”.
Fig. 6: Block diagram of alcohol detection.
If alcohol is detected then microcontroller will give
the signal to the L293D driver which will generate
a signal to the converter of the circuit. Also L293D
motor driver will control the operation of the DC
motor. Thus if alcohol is detected then bike will
never get start. Further using the GSM modem,
message will passed to the nearby control room or
any unique number.
Page 947
Imperial Journal of Interdisciplinary Research (IJIR)
Vol-2, Issue-11, 2016
ISSN: 2454-1362, http://www.onlinejournal.in
Module D: Obstacle Avoider Circuit.
In this module, we will design an obstacle avoider
wheelchair with AVR ATmega89S52
microcontroller using Analog IR Sensor. This
wheelchair is an automatic wheelchair i.e no
manual control is needed. This wheelchair can
detect obstacles from its three sides- front, left and
right side. The robot changes its direction of
movement, whenever it detects any obstacle in its
path from any side. Here, we will use three analog
IR sensors to detect obstacles from the three sides
of the wheelchair. The three sensors will be placed
in the front side, left side and right side of the
wheelchair pointing towards their respective side.
Whenever an obstacle comes in front of any sensor,
there will be a change in the output of that
particular analog IR sensor and this change will be
detected by the ATmega89S52 microcontroller.
But the outputs of the three analog IR sensors are
analog in nature, so these signals cannot be
processed directly by the microcontroller. For this,
we will use the ADC of the ATmega89S52
microcontroller to convert the analog signals to
digital values. After converting the analog signals
of analog IR sensors to digital values, the
ATmega89S52 microcontroller will compare the
sensor values with a reference value i.e threshold
value (3V in our case). According to the output of
the above comparisons, the ATmega89S52
microcontroller will send the required control
signal to the DC motor driver (L293D) of the
wheelchair to move the wheelchair in forward, left
or right direction. The above processes of ADC
conversions, sensors output comparison and robot
control continues forever till the power source is
there.
5 Conclusions:
Accelerometers have a secure place in the
movement of equipment based on actions done.
The system can be made free from challenges and
will be cost effective in the near future. Calibration
though at times is problem but with more
introspection and research better calibration and
performance can be achieved. The system
developed by us despite calibration errors and
problems still is able achieve accuracy of 88-95%,
further improvements can used to achieve an
accuracy of 95-99%.
The system proves a very competitive performance
computationally and in terms of recognition
accuracy. Interesting topic to research is the
problem of tilting. As mentioned earlier, tilting of
the remote can lead to erroneous recognition if not
taken into account. Therefore, in our proposed
system, subjects were requested to hold the remote
Imperial Journal of Interdisciplinary Research (IJIR)
in a natural way while performing the gestures and
to avoid any tilting of the remote as much as
possible. However, this way of holding the remote
can result in some inconvenience to users of the
system. Consequently, a system which is immune
to tilting of the accelerometer is definitely a
desirable one.
6 References:
[1]. CihanTopal, AtakanDoğan and
ÖmerNezihGerek, “A Wearable Head-Mounted
Sensor-Based Apparatus for Eye Tracking
Applications” VECIMS 2008 – IEEE International
Conference on Virtual Environments, Human-
Computer Interfaces, and Measurement Systems
Istambul, Turkey, 14-16 July 2008
[2]. Y. Chen, ―Application of Tilt Sensors in
Human–Computer Mouse Interface for People with
Disabilities‖, in IEEE Transactions Neural Systems
and Rehabilitation Engineering, Vol. 9, No. 3,
September 2001, pp. 289–295.
[3]. P. Jia, H. Hu, T. Lu and K. Yuan, “Head
Gesture Recognition for Hands-free Control of an
Intelligent Wheelchair”.
[4]. P.S. Gajwani and S.A. Chhabria. Eye motion
tracking for wheelchair control. International
Journal of Information Technology, 2(2):185– 187,
2010.
[5]. Ericka Janet Rechy-Ramirez, Huosheng Hu
and Klaus McDonald- Maier, “Head movements
based control of an intelligent wheelchair in an
indoor environment”, Proceeding of the 2012
IEEE, international conference on robotics and
biomimetic, December 11- 14, 2012, Guangzhou,
China.
[6]. Rabiner L. R., “A Tutorial on Hidden Markov
Models and Selected Applications in Speech
Recognition”, Proc.
[7].Moon, M. Lee, J. Chu, and M. Mun, "Wearable
EMG-based HCI for Electric-Powered Wheelchair
Users with Motor Disabilities," Proc. of the 2005
IEEE Int. Conf. on Robotics and Automation, pp.
2649-2654, 2005.
[8]. Rajesh KannanMegalingam, Ramesh
Nammily Nair, SaiManojPrakhya,” Automated
Voice based Home Navigation System for the
Elderly and the Physically
Challenged”proc.ICACT2011P.603-08,Feb.2011.
[9]. J.Z. Yi , Y.K. Tan Z.R. Ang, “Microcontroller
Based Voice- Activated Powered Wheelchair
Control”, ACM ublishers, 2007.
Page 948
Imperial Journal of Interdisciplinary Research (IJIR)
Vol-2, Issue-11, 2016
ISSN: 2454-1362, http://www.onlinejournal.in

[10]. A. AnandKumar “BasicFundamentals of

Digital Circuits” Second Ed. PHI Learning.

[11].http://www.rhydolabz.com/index.php?main_p
age=product_info&cP ath=124&products_id=1025

[12]. Muhammad Ali Mazidi “The 8051

Microcontroller and Embedded System” Second
Ed. Pearson Education.

[13]. M. J. Mathie, J. Basilakis and B. G. Celler,A

System For Monitoring Posture And Physical
Activity Using Accelerometers 2001 in
Proceedings of the 23rd Annual EMBS
International Conference(Istanbul, Turkey, October
25-28). 3

[14]. Paralysis resource centre ,Avalaible :

http://www.christopherreeve.org/site/c.mtKZKg
MWKwG/ b.5184189/ k.5587/
Paralysis_Facts__Fi gures.html

Imperial Journal of Interdisciplinary Research (IJIR) Page 949