Você está na página 1de 7

ISSN 2394-3777 (Print)

ISSN 2394-3785 (Online)


Available online at www.ijartet.com
International Journal
al of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue X, March 2015 in association with
HOLY GRACE ACADEMY OF ENGINEERING
ORGANIZES
NATIONAL LEVEL CONFERENCE ON INNOVATIVE ENGINEERING
TH
(16-20 MARCH 2015)

DIGITAL VOCALIZER
Delna Domini1, Sreejith s2, Aksa David3
B.Tech, Electronics and Communication, Holy Grace Academy of Engineering, Mala,
Mala, India
Asst. Professor, Electronics and Communication, Holy Grace Academy of Engineering, Mala
Mala, India
Asst. Professor, Electronics and Communication, Holy Grace Academy of Engineering, Mala
Mala, India

Abstract: This paper presents a gesture recognition model; known as Digital Vocalizer. Digital Vocalizer is an arduino UNO
based system and this system is designed to make the communication gap between dumb, deaf and blind communities and their
communication with thee normal people as less as possible. Here in this paper, we mentioned the people by normal people who
have the abilities to see, hear and talk. Deaf people make use of the sign languages or gestures to convey what he/she needs to
say; but it is impossible to understand by a person who can hear. So we come on a conclusion to make an easy prototype by
taking some of those gestures and convert them into visual and audio forms. So the communication gap can be reduced as much
as possible. For that we make use of arduino UNO Board as Atmega 328 Controller to interface all of the actuators and sensors.
Sensors are fixed on the palm and fingers. Those sensed values will give the information about the parameters like finger ben
bend,
hand position angle; and these are converted
nverted into electrical signal and switch to arduino UNO. Arduino UNO takes the actions
according to the gestures.

All Rights Reserved © 2015 IJARTET 676


ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal
al of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue X, March 2015 in association with
HOLY GRACE ACADEMY OF ENGINEERING
ORGANIZES
NATIONAL LEVEL CONFERENCE ON INNOVATIVE ENGINEERING
TH
(16-20 MARCH 2015)

values. Reference [14] explains a Nonspecific-User


Nonspecific Hand
Keywords: Include at least 4 keywords or phrases Gesture Recognition technology. An Algorithm and its
Application for 3D Interaction is discussed in reference [15].
I. INTRODUCTION
Communication is a crucial part of human beings life; the III. METHODOLOGY
lack of proper communication will cause severe problems. Block diagram of the whole system
ystem is shown Fig.1. The
The communication is mainly done by gestures and speech. system consists of:
So a complete coordination of these two is necessary. A  Data Glove
number of hardware techniques are used for gathering  Tilt Detection
information about body positioning; typically either image
image-  Bend Detection
based (using cameras, moving lights etc) or device
device-based
 Adriano Board
(using instrumented gloves, position trackers etc.), although
hybrids are beginning to come about [1], [2], [3].  Audio Processor
However, first we need to get the data; after that  LCD Display
we need to recognize those detected data from the glove and
it is considered as the second step. And the researches on
this are in progress. This research paper deals with the data
from a digital data glovee for the use of recognition of
gestures. This system converts these signs and into visual
and speech. The conversion of data to visual and speech is
achieved by the audio processor.

II. BACKGROUND
A lot of research works have been done in the field of Fig.1: Block Diagram of the System
gesture recognition; many are there in progprogress. The most
recent research; “Recognition of Hand Gestures Using Digital glove is consisted of two sensors. They are
Range Images”” is explained in reference [1]. Reference [2] flex sensors and accelerometer as tilt sensor. The tilt
gives a small frame work of the Hybrid Classifiers detection module and the bend detection module detect the
Classifiers.
Reference [3] gives a brief explanation off classification output of the accelerometer and flex sensor respectively.
respectively The
combinations of these results will lead to the generation of
some binary addresses; and those 8-bit
bit addresses are directly
switched to the audioo processor. For each gesture, the binary
address will be different.. Audio processor processes it and
Microcontroller and Sensors Based Gesture speaks. The data will be sent to the LCD which displays the
Gesture. “Microcontroller
Vocalizer” depicts a vocalizer by using 8051 word corresponding word.
microcontroller. In reference [4] a Survey on Gesture
Recognition has been done in reference [6]. Again a brief IV. SYSTEM DESCRIPTION
description about the Data Glove is there on reference [7].
The methods to improve the recognition are being framed in A. Data Glove
reference [8]. The requirements of a Self organized Flex lex sensors and accelerometer as tilt sensors are the sensors
recognition system are explained in reference [9]. A Free which are placed on the data glove.
glove The output of the
Hand Tracking technology is explained in reference [10]. A accelerometer is detected by the tilt detection module, while
Time Gesture Recognition system is discussed in the output of the flex sensors is detected by the bend
Real-Time
reference [11]. Reference [12] gives an analysis of Dynamic detection module. The combination of these results of two
Hand Gesture Recognition System. Reference [13] shows a sensors is given to the arduino UNO for further operations.
system with multiple sensors which will give more accurate

All Rights Reserved © 2015 IJARTET 677


ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal
al of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue X, March 2015 in association with
HOLY GRACE ACADEMY OF ENGINEERING
ORGANIZES
NATIONAL LEVEL CONFERENCE ON INNOVATIVE ENGINEERING
TH
(16-20 MARCH 2015)

This combined result will give the overall gesture of the


hand. As the amount of carbon increases the resistance
I: Bend sensor decreases. Circuit diagram of flex sensor is shown below in
‘Flex
Flex Sensor’ or ‘Bend Sensor’ is a sensor that changes its Fig.5.
resistance depending on the amount of bend on the sensor. In
this system, five flex sensors are there on the data glove
glove.
This are stitched on each of the finger of the hand glove and
they records the static and dynamic movements of the
fingers. Fig.2 shows how a flex sensor looks like. They are
usually in the form of a thin strip from 1”
1”-5” long that vary
in resistance range.

Fig.5: Circuit diagram of flex sensor

Fig.2: Flex Sensor The main applications of flex sensors are


Robotics, Automotive controls, Fitness Products,
Produ Medical
The Flex sensors are mainly made up of resistive Devices, Animatronics Virtual irtual Reality gaming consoles etc.
Carbon elements. Flex sensors have great form form-factor. The
sensor produces a resistance output according to the bend
radius. As the radius reduces, the resistance value increases.
They convert the change in bend into electrical resistance. II: TILT SENSOR
As the bend increases the resistance value also increases. In this system, Accelerometer is used to sense the tilting or
The Change in resistance with increasing bend is shown in slanting of the hand. An accelerometer is a device that can
Fig.3. measure dynamic acceleration (vibrations) and static
acceleration (gravity). Being that it measures this, it can
sense movements/lack-of-movements,
movements, tilts and even
rotation also. A microcontroller is required to read and
record the tilt. Here in this system, we are using arduino
microcontroller. The unit of the output of the accelerometer
is mg (Milligram). Hence the mapping of the readings is
necessary. Thereby we can display these readings on the
computer.
ADXL335 accelerometer is used in this digital
Fig.3: Change in resistance with increasing bend
vocalizer. The Output of accelerometer is analog and it
In simple words, flex sensors are analog resistors. ranges from 1.5 to 3.5 volts. The output of the accelerometer
They work as variable analog voltage dividers and it is is given to the arduino. ADXL335 interface with Arduino is
depicted in fig.4.

Fig.4: flex sensors as variable analog voltage dividers

All Rights Reserved © 2015 IJARTET 678


ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal
al of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue X, March 2015 in association with
HOLY GRACE ACADEMY OF ENGINEERING
ORGANIZES
NATIONAL LEVEL CONFERENCE ON INNOVATIVE ENGINEERING
TH
(16-20 MARCH 2015)

shown below in Fig.6.

Fig.6: arduino UNO

Each flex sensors’ output is given to the separate analog pins


of arduino board. At first the output of these five flex
Fig.6: ADXL335 Interface with Arduino UNO sensors are being evaluated and pulse width is calculated,
after the evaluation of all the five flex sensors they are
Six pins are there on the ADXL335 accelerometer. It is combined with the values of significant tilt /slant. Flex
shown in table.1. Two pins are there for power supply and sensor is very sensitive. Even a small bend is monitored it
ground, X, Y and Z pins are the analog pins. The last pin will give infinite levels of bends. Hence the values of
known as test pin can be left as disconnected. First power bending are quantized into ten levels. This makes easier to
supply must be up then wire up the analog pins to the analog determine how much bending of the finger has happened.
inputs of microcontroller. Since the outputs of both the sensors are analog,
TABLE I this is not measurable by the microcontroller. It is necessary
PIN DESCRIPTION OF ADXL335 to convert it into digital form. A lot of Analog to digital
converter ICs (ADC) are available in the market. However,
in this system, we are using arduino UNO which has inbuilt
ADC. These two digital values are being evaluated in the
further more steps.
As a continuation of the above steps, the arduino
checks the digital output fromm the ADC. The significance of
the data which is received from the ADC’s is checked by the
Arduino.. Significant means that hand is delivering some
meaningful gesture. The microcontroller compares the
output of the ADC with the already stored values. According
to this comparison, the arduino determines that, whether the
B. Gesture Detection gesture is significant or not.If it is a significant gesture, the
Gesture recognition is handled by arduino UNO. The main microcontroller moves on to the next step; the eight bit
function of arduino UNO is to compare the detected values binary data to the audio processor and LCD will be sent.
by the sensors with the predefined values. According to that
comparison it sends some binary addresses to the LCD and C. Audio Processor
the Audio processor. The arduino UNO would have two Audio processor speaks out against the significant gesture.
input sections; they are from the flex sensors and from the Here we are using aPR33A3. It iss capable of producing 8
accelerometer. voice messages. For that the voice messages must be
recorded. To record these messages, a microphone can be
used. A lot of microphones are available in the market. Here
in this system an electret microphone is used.
I. ELECTRET MICROPHONE

All Rights Reserved © 2015 IJARTET 679


ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal
al of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue X, March 2015 in association with
HOLY GRACE ACADEMY OF ENGINEERING
ORGANIZES
NATIONAL LEVEL CONFERENCE ON INNOVATIVE ENGINEERING
TH
(16-20 MARCH 2015)

A microphone is a transducer which converts sound energy If your supply is 5v, you need to bias the output to
to electrical signals. Transducers are devices which convert 2.5v to get maximum swing from your input signal. So a
energy from one form to other. Its working principle is pre-amplification
amplification can be used along with microphone signal
opposite to a speaker. They are available in different shapes in order to activate the analog input of arduino and it is
and sizes. Commonly used microphone is electret condenser shown in Fig.13. The output waveform of the microphone
microphone. It is shown in Fig.9. It is used in mobile with the pre-amplification
tion circuit is shown in Fig.14.
phones, laptops, etc. It is mainly used to detect minor
sounds or air vibrations. The top of the electret microphone
is covered by a porous material. It filters out the dust
particles.

Fig.9: Electret Microphone


.
Fig.13:Pre-amplification
amplification circuit for Electret Microphone
The two legs as shown in fig.9 are used to make
electrical connection with the circuit. The sound signals/air
vibrations pass through the porous material and falls on the
diaphragm through the hole at the centre. This change in
capacitance due to the presence of sound waves produces
variance in voltage and sends corresponding
orresponding el
electrical
signals. The frequency response of electret microphone is
shown in Fig.11. Test circuit for microphone and amplifier
modules is shown in Fig.12

Fig.14: The output waveform of microphone with pre-amplification circuit

Arduino
ino generates the 8 bit addresses according to
the data from the bend and tilt detection modules. Now the
arduino can understand the meaning of each gestures and it
is capable to send thee corresponding message address.
address As we
Fig.11: Frequency
requency response curve of electrets microphone need to generate the voices to each significant gesture, we
switch these 8-bit addresses to the audio processor IC.
IC Here
we are using aPR33A3 audio processor.
processor

Fig.12: Test circuit for microphone and amplifier modules

All Rights Reserved © 2015 IJARTET 680


ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal
al of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue X, March 2015 in association with
HOLY GRACE ACADEMY OF ENGINEERING
ORGANIZES
NATIONAL LEVEL CONFERENCE ON INNOVATIVE ENGINEERING
TH
(16-20 MARCH 2015)

module. An eight bit address against each significant gesture


is sent to LCD module. This eight bit address decides what
should be displayed on the LCD.

V. Overview of the System

The Fig.17
17 depicts the over view of the Digital Vocalizer.

Fig.15:aPR33A3 Interface with Arduino UNO


Some particular allophones are allocated to each
word and for those allophones some addresses are defined in
audio processor IC. These 8 bit addresses es are sent to the
aPR33A3 to speak those particular wordss. The summary is Fig.17: An overview of the whole system
that it is necessary to know the 8 bit digital address
addresses of each 6. Conclusion and Future Enhancements This paper gives
word or sentence. The microcontroller sends the these addresses a detailed structure of a system is designed to make the
to aPR33A3. The address will locate the allophone of the communication gap between dumb, blind and deaf
word. The aPR33A3 gives an output signal and it can be communities and their communication with the normal
amplified to make it louder. The speaker takes the ooutput of people who can speak, see and talk as less as possible. The
this amplifier as the input and it speaks the voice message
messages sign language is being by the dump communities used for
corresponding to the 8-bit addresses. their communication. But these sign languages cannot be
used for the communication of dump communities with the
D. LCD Display blind as well as normal people. Here we can use this Digital
Vocalizer which will generate voice and display according
The audio processor gives a provision for the to these signs. This display will help deaf communities to
communication of dumb with the normal people and with reduce the communication gap also.
the blind people as well. But the he communication gap The future enhancements:
between the dump people and the deaf people cannot be 1. A provision for ZigBee standard for “Digital
reduced with the help of audio processor IC
IC. To avoid this Vocalizer”.
condition a display can be attached along with the audio 2. A new design with TTS256t to improve message
processor. storage capacity.
3. Accurate monitoring of the static dynamic
movements involved in “Digital Vocalizer”.
4. A system with whole jacket, which could be
capable of conveying the movements of animals.
5. The replacement of virtual reality application like
joy sticks with the “Digital Vocalizer” (Data
Glove).
Fig.16: LCD Interface with Arduino UNO
VI. REFERENCES
JHD162A is the LCD module used here. JHD162A is a 16×2
LCD module based on the HD44780 driver from Hitachi. [1] Kazunori Umeda Isao Furusawa and Shinya
The JHD162A has 16 pins and can be operated in 44-bit mode Tanaka, “Recognition of Hand d Gestures Using
Range Images”, Proceedings of the 1998 IEEW/RSJ
or 8-bit mode. The microcontroller has already International Conference on intelligent Robots and
generated the 8 bit address bit and it is bbeing sent to the Systems Victoria, B.C.,
., Canada October 1998, pp.
audio processor, the same address is being sent to the LCD 1727-1732

All Rights Reserved © 2015 IJARTET 681


ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal
al of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue X, March 2015 in association with
HOLY GRACE ACADEMY OF ENGINEERING
ORGANIZES
NATIONAL LEVEL CONFERENCE ON INNOVATIVE ENGINEERING
TH
(16-20 MARCH 2015)

[2] Srinivas Gutta, Jeffrey Huang, Ibrahim F. Imam, Department of ECE,ICICES2014 - S.A.Engineering College,
and Harry Wechsler, “Face and Hand Gesture Chennai, Tamil Nadu, India
Recognition Using Hybrid Classifiers”, ISBN: 00- [15] lianfeng Liu, Zhigeng Pan, Xiangcheng Li "An Accelerometer-
Accelerometer
8186-7713-9/96, pp.164-169 Based
[3] Jean-Christophe
Christophe Lementec and Peter Bajcsy, Gesture Recognition Algorithm and its Application for 3D
“Recognition of Am Gestures Using Multiple Interaction";
Orientation Sensors: Gesture Classification”, 2004 ComSIS Vol. 7, No. I, Special Issue, February 2010.
IEEE Intelligent Transportation Systems Conference
Washington, D.C., USA, October 34, 2004, pp.965
pp.965-
970
[4] Microcontroller and Sensors Based Gesture Vocalizer
Proceedings of the 7th WSEAS International Conference on
SIGNAL PROCESSING, ROBOTICS and AUTOMATION
(ISPRA '08), University of Cambridge,
ridge, UK, February 20
20-22,
2008
[5] Sushmita Mitra and Tinku Acharya,”Gesture
Recognition: A Survey”, IEEE TRANSACTIONS ON
SYSTEMS, MAN, AND CYBERNETICS
CYBERNETICS—PART C:
APPLICATIONS AND REVIEWS, VOL. 37, NO. 3,
MAY 2007, pp. 311-324
[6] Recognition of a Hand-Gesture Based on
Self-organization Using a Data Glove, 0-7803
7803-5871-
~/99/$10.080 1999 IEEE

[7] Sanshzar Kettebekov, Mohammed Yeasin and


Rajeev Sharma, “Improving Continuous Gesture
Recognition with Spoken Prosody”, ”, Proceedings of the
2003 IEEE Computer Society Conference oon Computer
Vision and Pattern Recognition (CVPR’03), ISBN #
1063-6919/03, pp.1-6
[8] Masumi Ishikawa and Hiroko Matsumura,
“Recognition of a Hand-Gesture
Gesture Based on Selforganization
Using a Data Glove”, ISBN # 0-7803-
5871-6/99, pp. 739-745
[9] Hand Gesture Recognitionion Using Accelerometer
Sensor for Traffic Light Control System.
Shirke Swapnali, Student of ENTC Dept., SKNCOE,PUNE
Pune,India
[10] Gloved and Free Hand Tracking based Hand Gesture
Recognition,
978-1-4673-5250-5/13/$31.00
5/13/$31.00 ©2013 IEEE
IEEE, ICETACS 2013
[11] Toshiyuki Kirishima, Kosuke Sato and Kunihiro
Chihara, “Real-Time
Time Gesture Recognition by Learning
and Selective Control of Visual Interest Points”, IEEE
TRANSACTIONS ON PATTERN ANALYSIS AND
MACHINE INTELLIGENCE, VOL. 27, NO. 3,
MARCH 2005, pp. 351-364
[12] Attila Licsár and Tamás Szirány, “Dynamic
Training of Hand Gesture Recognition System”,
Proceedings of the 17th International Conference on
Pattern Recognition (ICPR’04), ISBN # 1051
1051-
4651/04,
[13] A Method of Hand Gesture Recognition based on
Multiple Sensors, Fan Wei, Chen Xiang, Wang Wen Wen-hui, Zhang
Xu ,Yang Ji-hai,978-1-4244-4713-8/10/$25.00
8/10/$25.00 ©2010 IEEE
[14] Nonspecific-User
User Hand Gesture Recognition By
Using MEMS Accelerometer, Jayaraman D
Student Member of IEEE

All Rights Reserved © 2015 IJARTET 682

Você também pode gostar