Gesture Recognition Based Device Control Using MEMS Accelerometer

Full text

Received: 07 August, 2017; Revised: 14 October, 2017; Accepted: 15 November, 2017
*Corresponding Author DOI: 10.26674/ijesacbt/2017/49248
International Journal of Engineering Science, Advanced Computing and Bio-Technology
Vol. 8, No. 4, October – December 2017, pp. 256 - 264
Gesture Recognition Based Device Control Using
MEMS Accelerometer
*Kalyan Kasturi1, K. Sri Mourya2, I. Mahendra2 and Vikas Maheshwari1
1 Bharat Institute of Engineering and Technology, Hyderabad – 501510, India
2R.M.D. Engineering College, Chennai-601206, India
Email: kalyankasturi@hotmail.com
Abstract: Gesture recognition is one form of human interaction that can be used effectively in human-
computer- interaction. Human-computer-interaction is related to exchange of information between
human beings and the computers. In this research work by recognizing the particular gesture given by the
user, the microcontroller activates a corresponding relay circuit which in turn, operates the appropriate
device. The microcontroller was programmed to transmit the status of a particular device, whether ON
or OFF, to a laptop using the ZigBee wireless protocol. Gesture recognition provides a user-friendly modus
operandi and a personalized touch to the operation of household devices. We used MEMS Accelerometer
to capture motion trajectory information based on accelerations of objects to recognize gestures. The
Accelerometer sensor measures acceleration values related to the gesture movements and passes this
information to a PIC microcontroller, which will operate the corresponding device. Other authentication
methods such as voice activated command system require very high accuracy. Especially when used by
stressed or distracted people, the accuracy of authentication by voice command system is lower, whereas
higher accuracy of device authentication can be obtained by the proposed gesture recognition. The
proposed method is also highly cost effective compared to the voice command authentication systems.
Keywords: MEMS Accelerometer, Gesture Recognition, Device Control
1. Introduction
MEMS stands for Micro-Electro-Mechanical systems. The term ‘MEMS’ or Micro-
Electro-Mechanical-Systems was first used by R. Howe as well as several other researchers
[1] to differentiate the mechanical elements developed at microelectronic circuit scale than
the mechanical elements developed using lathe machining. The MEMS devices have features
below 10 micro-meters that are developed using micro fabrication technology. The MEMS
devices can be manufactured in silicon, polymer, quartz and metals.
MEMS have been used in diverse applications, from display technologies to sensor
systems to optical networks. MEMS are attractive for many applications because of their
small size and weight, which allow systems to be miniaturized. The MEMS technology
allows the development of devices to operate or function in the micro-world, where it is not
feasible to use ordinary machines. Another advantage of MEMS technology is that it
provides the possibility of miniaturization of machines for cost effective operation and
better precision [2].
To perform new development of MEMS devices, it is better to start from stable MEMS

International Journal of Engineering Science, Advanced Computing and Bio-Technology
257
technology platforms. IMEC platforms provide means to perform MEMS integration, RF-
MEMS, MEMS interconnection and packaging. In the paper by Pieters [3], the use of
IMEC's CMORE for technology development, prototyping and small volume production is
described. In the recent years MEMS technology is used for manufacturing accelerometers.
These MEMS accelerometers are small-sized devices with simple operating procedure. A
basic MEMS accelerometer can be a cantilever beam with a proof mass. Due to the external
accelerations the proof mass is displaced from its neutral position and this displacement can
be measured by the calculation of capacitance difference [4].
Gesture recognition has been investigated by several researchers. In the research study
by Cao and Balakrishnan [5], a plastic stick was tracked in 3-dimensional space, for use as
an input device for large scale displays. The endpoints of the stick were tracked by a pair of
cameras. By using computer vision techniques, the endpoints of the stick were tracked in
three dimensions. A rich vocabulary of actions was encoded by the endpoints of the stick.
In the research study by Kela et al. [6] gesture control using accelerometer was investigated
as an alternative interaction method. They developed gesture commands that can be trained
by the user to perform the operations on external devices. They conducted an investigation
to compare the utility of gesture based control with other methods such as speech, laser pen
and RFID objects. The results of their investigation also showed that gesture control was a
preferred choice for operations related to spatial association in design environment control.
Liu et al. [7] developed an algorithm for performing interactions based on gestures using
accelerometers. They developed ‘u-Wave’ algorithm for identifying various gestures using
accelerometer readings. They reported to achieve an accuracy of 98.6%.
The research work by Meenaakumari. M and M. Muthulakshmi [8] investigated the
development of hardware module consisting of a MEMS accelerometer, microcontroller,
and Zigbee wireless transmission module for sensing and collecting accelerations of
handwriting and hand gesture trajectories. Another research paper by S.Karthick and
R.Radhika Shridevi [9], investigated the development of an automatic gesture recognition
algorithm to identify individual gestures in a sequence. The particular gesture was
recognized by comparing the acceleration values with the stored templates that has been
stored in EEPROM.
2. System Model for Device Control
In this research paper, we used MEMS based accelerometer to perform gesture
recognition to perform the controlling function of various household devices in an orderly
manner. The accelerometer is connected to the PIC microcontroller which performs gesture
recognition.
The output of the PIC microcontroller is in turn connected to three relay devices as
shown in Figure 1. Based on the particular gesture, one of the three relays is activated. Each
of the three relays is connected to a household device such as a lamp, an LED light and a

Gesture Recognition Based Device Control Using MEMS Accelerometer 258
fan. The activation of a particular relay as determined by the particular gesture, eventually
turns on a particular device and facilitates device monitoring.
Figure 1: System model of the household device control system.
3. Components Of Device Control System
In this research work we have selected the MEMS accelerometer developed by Analog
Devices. The accelerometer ADXL325 used in this research work is a low-power, triple-axis
accelerometer. It can measure acceleration with a full-scale range of ±5g. It can be used to
measure both dynamic acceleration and static acceleration. Since this accelerometer operates
using low power and uses a voltage range of 2.4 Volts to 5.25 Volts, it is suited for battery
powered applications. This accelerometer is connected to the PIC microcontroller using an
analog-to-digital converter.
The acronym ‘PIC’ stands for programmable intelligent computer. PIC
microcontrollers are easily programmable and easy to interface with other devices. The PIC
microcontrollers consist of a microprocessor, memories, I/O ports, timers, and other
hardware components. In this research paper we used the PIC microcontroller PIC16f877.
This microcontroller is a high performance RISC processor. It uses reduced instruction set
for achieving higher processing speed, it supports 35 instructions. All the instructions are
single-cycle instructions except for the branch instructions which are two-cycle instructions.
The microcontroller has an operating speed of 200 MHz. This microcontroller was
connected to a 10-bit analog-to-digital converter (ADC). The data memory of the
MICRO
CONTROLLER
RELAY
1
RELAY
2
RELAY
3
LAMP
(1)
LED
(2)
FAN
(3)
MEMS
ACCELERO
METER

International Journal of Engineering Science, Advanced Computing and Bio-Technology
259
microcontroller is in 8-bit byte format. Hence the 10-bit analog-to-digital converter (ADC)
output was scaled using left shifting into the 8-bit format.
The microcontroller supports 8 KB flash memory which was used to store the program
for gesture recognition. The gesture recognition program used in this research work can
identify 3 different gestures. We used 3 simple gestures based on the letter C. The 3 gestures
are used in this research work are normal C, reverse C and inverted C. Based on the
particular gesture identified by the microcontroller, a particular relay circuit of the 3
available relay circuits was activated.
PIC microcontrollers are highly popular microcontrollers developed by Microchip
Technology. So far more than 12 billion of these devices have been used in various designs
of embedded system applications.
There are 3 relay circuits that are used in this research work. Each of the 3 relays is
connected to a particular I/O port of the microcontroller. When the microcontroller
identifies a particular gesture, it activates the corresponding I/O pin and the particular relay
is turned on.
A relay is a switch which can be operated by an electrical means. A relay is used when
several components must be controlled by using some type of electrical signal. A simple
electromagnetic relay consists of an iron core, armature and moving contacts. The armature
is hinged to the iron core and mechanically connected to the moving contacts. A coil is
wound around the armature. When an electrical current passes through the coil, it generates
a magnetic field that activates the armature and the consequent movement of the movable
contacts makes a connection with a fixed contact.
A relay is an electrically operated switch. Many relays use an electromagnet to
mechanically operate a switch, but other operating principles are also used, such as solid-
state relays. Relays are used where it is necessary to control a circuit by a low-power signal
(with complete electrical isolation between control and controlled circuits), or where several
circuits must be controlled by one signal.
The first relays were used in long distance telegraph circuits as amplifiers: they repeated
the signal coming in from one circuit and re-transmitted it on another circuit. Relays were
used extensively in telephone exchanges and early computers to perform logical operations.
A type of relay that can handle the high power required to directly control an electric motor
or other loads is called a contactor. Solid-state relays control power circuits with no moving
parts, instead using a semiconductor device to perform switching. Relays with calibrated
operating characteristics and sometimes multiple operating coils are used to protect
electrical circuits from overload or faults; in modern electric power systems these functions
are performed by digital instruments still called "protective relays".
The status of the various devices (whether ON or OFF) is transmitted from the
microcontroller to a laptop or personal computer using ZigBee wireless protocol.
ZigBee is a specification for low-power wireless transmission up to a distance of 10 meters.
It is based on IEEE 802.15.4 specification and used for personal area networking (PAN).

Gesture Recognition Based Device Control Using MEMS Accelerometer
260
Since ZigBee can span the area of a house and is less expensive, it is used for device
monitoring in this research work. The microcontroller is connected to the ZigBee transmit
section via serial port interface.
If for example device 1 is on, then a message ‘ON 1’ will be transmitted from the
microcontroller to the ZigBee transmit section. Using wireless transmission via ZigBee
protocol, this message is in turn transmitted from the ZigBee transmit section to the ZigBee
receiver section which is connected to the laptop or personal computer and the message will
be displayed on the laptop or computer.
4. Demonstration Kit
Figure 2 illustrates the demonstration kit showing the important hardware parts of the
device control system. The relay part consisting of three relays such as Realy1, Relay2 and
Relay3 is depicted on the left of the Figure 2. The relays are located between the PIC
microcontroller and the three devices such as Device1, Device2 and Device3. The relays
provide the actuating signals to turn on the devices.
The PIC microcontroller is located in the center of the Figure 2. The microcontroller
performs the computing tasks using the accelerometer input and generates the actuating
signals. Most of the programming was performed using MPLAB.
Figure 2: Demonstration kit of the device control system.

International Journal of Engineering Science, Advanced Computing and Bio-Technology
261
MPLAB is a software integrated development environment for the development of
embedded applications on PIC microcontrollers. Microchip Technology has developed
MPLAB software.
MPLAB is designed to work with MPLAB-certified devices such as the MPLAB ICD 3
and MPLAB REAL ICE, for programming and debugging PIC microcontrollers using a
personal computer. PIC Kit programmers are also supported by MPLAB.
In Figure 2, the first electronic system to the right, is the interface containing the
accelerometer ADXL325.
5. Experimental Results
When a particular device is turned on by the accelerometer gesture activation, the
corresponding device number is displayed on the laptop connected to the demonstration
kit. The device number 1 is used for Lamp. The device number 21 is used for LED. The
device number 3 is used for Fan. A sequence of the device monitoring status is presented in
Figure 3.
Figure 3: The ON/OFF status of various devices connected to the device control system.
The operation of the demonstration kit for household device monitoring was verified
using the Putty software which is a free and open-source terminal emulator, serial console
and network file transfer application. It supports several network protocols, including SCP,
SSH, Telnet, rlogin, and raw socket connection. It can also connect to a serial port.

Gesture Recognition Based Device Control Using MEMS Accelerometer
262
When Lamp which is device 1 is activated then the display is “ON1”. If the LED which
is device 2 is activated then the display is “ON2”. Finally if the Fan which is device 3 is
activated then the display is “ON3”. When all the devices are not activated then the display
is “ALL OFF”.
It can be observed that in the beginning all the devices were inactive as depicted by the
“ALL OFF” tag. Next the device 3 was turned on as depicted by the “ON3” tag. After this,
all the devices were off are depicted by “ALL OFF” tag and then the device 2 was turned on
as depicted by the “ON2” tag. Finally the device 1 was turned on as depicted by the “ON1”
tag. The intermittent ‘ALL OFF’ tag indicates the temporary OFF status of all the devices.
6. Comparative Study And Discussion
In the research paper by V. Sundara Siva Kumar [10], a MEMS accelerometer based
system was simulated to enable the movements of the wheel chair of physically handicapped
individuals. They used Keil software for compilation part and Proteus 7 software for
simulation part. When a change in the hand gesture happened, then the MEMS
accelerometer generates a particular analog signal from mechanical signal. This is given as
input to the analog-to-digital converter (ADC) which converts the analog signal to digital
signal. According to the change in direction in the MEMS sensor, the micro controller
controls the motor direction by motor driver, either as right or left or forward or backward.
For future improvements, by combining the research work presented in this paper with the
research work given in [10], and using a larger set of gestures, an improved version of the
wheel chair movements can be implemented. By using a larger set of 32 gestures, the motor
connected to the wheel chair can be driven in 32 different ways and better control of the
wheel chair by the physically handicapped person can be achieved. Over the last decade,
IOT based systems for networking hand-held devices are becoming very popular [11]. By
integrating the IOT concepts with this research work, an IOT network for helping physically
handicapped individuals can be developed.
7. Conclusion
This paper investigated the use of MEMS accelerometer for household device control
via gesture recognition by a microcontroller. To capture the gestures a triple-axis
accelerometer ADXL325 was used.
The microcontroller PIC16F877 was used to identify three different gestures via the
software development environment MPLAB. Experiments with the demonstration kit
verified that a particular household device was activated based on the particular gesture.
Gesture based monitoring of household devices using MEMS accelerometer is one of the
technological steps towards building user-friendly and personalized technology for
consumers using gesture recognition.

International Journal of Engineering Science, Advanced Computing and Bio-Technology
263
The status of a particular device, whether ON or OFF, was transmitted to a laptop
using ZigBee wireless protocol, for monitoring purposes. The sequence order of the various
devices being turned ON was verified.
Further improvements can be made by increasing the number of gestures recognized
to 32 in order to control more number of devices. Another future scope of this research
work is to increase the number of recognizable gestures to 32 and develop an advanced
wheel chair with high resolution movements for the physically handicapped people.
References:
[1] N. Maluf, An introduction to microelectromechanical systems engineering, 1st edition, Artech House, Boston,
(1999).
[2] S. E. Lyshevski, Mems and Nems: systems, devices and structures, CRC Press, USA, (2002).
[3] P. Pieters, Versatile MEMS and MEMS integration technology platforms for cost effective MEMS
development, European Microelectronics and Packaging Conference, (2009), pp. 1-5.
[4] Lee, G. H. Yoon, J. Park, S. Seok, K. Chun and K. Lee, Development and analysis of the vertical capacitive
accelerometer, Sensors and Actuators, Elsevier Publication, 119(2005), pp. 8–18.
[5] X. Cao and R. Balakrishnan, VisionWand: interaction techniques for large displays using a passive wand
tracked in 3D, in Proc. ACM Symp. User Interface Software and Technology (UIST). Vancouver, Canada:
ACM, (2003).
[6] J. Kela, P. Korpipää, J. Mäntyjärvi, S. Kallio, G. Savino, L. Jozzo, and D. Marca, Accelerometer-based gesture
control for a design environment, Personal Ubiquitous Computing, 10(2006), pp. 285-299.
[7] J. Liu, L. Zhong, J. Wickramasuriya, and V. Vasudevan, uWave: Accelerometer-based personalized gesture
recognition and its applications, Pervasive and Mobile Computing, 5(2009), pp. 657-675.
[8] M. Meenaakumari and M. Muthulakshmi, MEMS Accelerometer Based Hand Gesture Recognition,
International Journal of Advanced Research in Computer Engineering & Technology, 2(2013), pp. 1886-1892.
[9] S.Karthick and R.Radhika Shridevi, Hand Gesture And Common Interaction Recognition Using MEMS,
International Journal of Advanced Research in Electrical, Electronics and Instrumentation Engineering,
3(2014), pp. 12252- 12258.
[10] V. Sundara Siva Kumar, G. Ramesh and P. Nagesh, MEMS based Hand Gesture Wheel Chair Movement
Control for Disable Persons, International Journal of Current Engineering and Technology, 5(2015), pp. 1774-
1776.
[11] K. Kasturi, P. Vishal Reddy, N. Achyuth Rao and S. Vinod, A Review of Architecture and Applications for
Internet of Things, Advances in Natural and Applied Sciences, 10(2016), pp. 402- 411.

Gesture Recognition Based Device Control Using MEMS Accelerometer 264
Authors’ Profile:
Kalyan Kasturi obtained B. TECH. degree in Electronics and Communication Engineering,
from V.R. Siddhartha Engineering College, Vijayawada, India in the year 1999. He received
M. S. and Ph.D. degrees from University of Texas at Dallas, USA in the years 2002 and 2006
respectively. Presently, he is attached with Bharat Institute of Engineering and Technology,
Hyderabad, India, as an Associate Professor in the Department of Electronics and
Communication Engineering. His research interests include DSP and Embedded Systems. He
has published numerous papers in International Journals and conferences. He is a recipient of
Fast Track Young Scientist Research Grant from Department of Science and Technology (DST).

Vikas Maheshwari passed B.Tech. degree in Electronics and Communication from U.P.
Technical University, Lucknow, U.P., India in the year 2006. He received the M.Tech. degree
in Microelectronics and VLSI from National Institute of Technology, Durgapur, West- Bengal
, India in the year 2010. He received PhD degree in Microelectronics & VLSI from National
Institute of Technology, Durgapur, West-Bengal, India in the year 2014. His research interest
includes Analog VLSI Design, VLSI interconnect modeling and optimization and Antenna
Design & optimization. He has published 35 International Journals and 45 International Conference papers.