Bq4301381388

B. Mamatha Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 3( Version 1), March 2014, pp.381-388
www.ijera.com 381 | P a g e
Transmission by an Embedded System with Enhancements in
Voice Processing Technologies
G.Sitha Annapurna1
,B.Mamatha2
, P.Annapurna3
,
1
B.Tech, M.Tech Assistant Professor in ECE Dept., Sri vasavi Institute of Engineering and Technology,
Nandamuru, Pedana (MD),Krishna D.t,A.P
2
B.Tech, M.Tech Assistant Professor in ECE Dept., Sri vasavi Institute of Engineering and Technology,
3
B.Tech,(M.Tech) Assistant Professor in ECE Dept., Sri vasavi Institute of Engineering and Technology,
Abstract
The paper reports that the robot can transmit the data such as video, audio, images. The robot can be controlled
using the human voice. There are two embedded systems first one is robot controlling system(MASTER) which
is used to control the robot, second one is voice controlled robot(SLAVE) which responds according to the
instructions coming from the controlling system. These two embedded systems are communicated through
wireless. We can use anyone one of wireless protocols such as IR, NFC, Bluetooth, Zigbee, WI-FI in order to
establish a bridge between the MASTER and SLAVE. The voice controlled robot can understand the
instructions with the help of the voice recognition system. Spinx-4 is a speech recognizer system written entirely
in the java programming language.Sphinx-4 started out as a port of Sphinx-3 to the Java programming language,
but evolved into a recognizer designed to be much more flexible than Sphinx-3, thus becoming an excellent
platform for speech research.Sphinx-4 is an HMM-based speech recognizer. HMM stands for Hidden Markov
Models, Sphinx-4 are a type of statistical model In HMM based speech recognizers.
Index Terms – Embedded system; voice processing; wireless transmission
I. Introduction
There have been considerable efforts to
achieve the hand-free control of computers. There are
usually two ways to control computers and robots on
a hand free basis voice commands and gesture
commands. These hand-free input methods have been
considered as an alternative to the traditional
keyboard inputs. The main objective of this paper is
to implement a voice recognition system onto a robot
for controlling its movement with a high degree of
accuracy and robustness. The users will be able to
control the robot by either speaking directly to the
robot or speaking to a microphone connected to a
computer, which will extract the command and pass
it to robot through wireless network connection. It is
also possible to transmit the data from voice control
robot to controlling system. The data which is sent by
robot is received by personal computer. We can see
on the display of the personal computer whatever
sent by the voice control robot.
II. Wireless protocol
2.1. Introduction
Bluetooth's primary function is to replace
cables. It's a technology built around short-range
radio links between mobile PCs, mobile phones and
other portable devices. The technology promises to
eliminate the confusion of cables, connectors and
protocols confounding communications between
today's high tech products.
2.2. Specifications:
The Following specifications are required
2.2.1 Open specification:
The Bluetooth Special Interest Group (SIG)
has produced a specification for Bluetooth wireless
communication that is publicly available and royalty
free.
2.2.3 Short-range wireless:
Today, much of the communication takes
place over cables. There are a wide variety of
connectors with many combinations of shapes, sizes
and number of pins. These cables can become quite
burdensome to users. With Bluetooth technology,
these devices can communicate without wires over
air using radio waves to transmit and receive data.
Bluetooth wireless technology is specifically
designed for short-range (nominally 10 meters)
communications. one result of this design is very low
power consumption, making the technology well
suited for use with small, handheld devices that
typically are powered by batteries.
RESEARCH ARTICLE OPEN ACCESS

2.2.4Voice and data:
Voice is now commonly transmitted and
stored in digital formats. Bluetooth wireless
communication makes provisions for both voice and
data, and thus it is an ideal technology for enabling
all sorts of devices to communicate using either or
both of these content types.
2.2.5 Anywhere in the world:
Many forms of wireless communications are
regulated in many parts of the world; radio frequency
spectrum usage often requires a license with strict
transmission power obligations. Bluetooth wireless
communications operate within a chosen frequency
spectrum that is unlicensed throughout the world (2.4
GHz). Thus, devices that employ Bluetooth wireless
communication can be used unmodified, no matter
where a person might be.
2.3. History
Bluetooth was invented in 1994 by L. M.
Ericsson in Sweden. Japp Hartseen is one of the co-
inventors to Bluetooth; he says that the technology
came about almost by accident. The original intention
was to make a wireless connection between
something like an earphone and a cordless headset
and the mobile phone. When they realized that they
could tap into a low radio frequency that required no
licensing and was available to anyone in the world
that wanted it, Hartseen and his colleagues at
Ericsson began to experiment with different
computer chips. They developed small radio chips
that made wireless connection between devices
containing them. To encourage adoption of Bluetooth
and with hope of developing a global standard,
Ericsson decided to give the specification for the
technology away for free. When the technique was
out there, The Bluetooth Special Interest Group (SIG)
was formed. The members were Toshiba, IBM,
Ericsson, Nokia and Intel. Today SIG have over 1200
company members.
As VLSI technology advances and
computing power grow in the past twenty years,
robots became more and more intelligent, robust and
consumed less and less power. Moreover, they are
required to handle more and more so called
teamwork. It means that they must be developed to
possess the capability of constructing a network and
performing cooperative works. A key driving force in
the development of cooperative mobile robotic
system is their potential for reducing the need for
human presence in dangerous applications [1]. It is
possible to establish a point to point connection using
Zigbee protocol. Instead of point to point connection,
In order To communicate with more than one robot
we use a Bluetooth Wi-Fi integrated chip. So that it is
possible to control the robots by selecting the
appropriate robots.[2-3]
III. Wireless data transmission
3.1 HARDWARE SYSTEM:
This system mainly adopts B/S structure,
embedded WEB server as the center, and captures the
video by USB camera, then carries on MPEG-4
compression. The compressed MPEG-4 frame RTP
package is transmitted through RTP protocol. As a
result, the user can see the video image directly that
come from WEB server through the IE browser in the
monitoring terminal. Figure.1 shows the hardware
construction of the system for remote video capture
and transmission based on ARM. We select the
ARM9 development board, provided by Beijing
Hengfeng Ruike Technology Co., Ltd, as the
hardware platform of the system. This system bases
on embedded chip S3C2410AL constructed on
ARM9, its basic frequency can be up to 203MHz,
and the board carries 64MB SDRAM and 64MB
NANDFLASH. The resources of the main board
mainly include: two USB host interface, a USB
device interface, a 10M/100M Ethernet port, a serial
port, a JTAG interface, touch screen and so on.
Figure.1 Hardware construction of the system for
remote video capture and transmission based on
ARM
3.2 SOFTWARE SYSTEM:
The platform of software system is
embedded operating system whose kernel is Linux-
2.6.30.4. Embedded Linux operating system has quite
a lot of advantages. The source code of operating
system opens completely, can be cut out and has low
cost; the kernel is small and stable, which of function
is powerful; the operating system supports various
hardware platform and multitasking processe, which
can provide better performance of real time. These
characteristics are very suitable for embedded
application, and Linux itself provides complete
TCP/IP protocol directly, which can be easily applied
to network.
3.2.1 A. Module of Embedded Web Server
In numerous Web server, there are many
lightweights, such as boaweb, thttp, httpd, etc.
however, if you want to get higher security and
provide convenient embedded Web server for the last
Web development period, those servers cannot meet
the case. So we select the recently released
AppWeb3.0, which has a high level security, to
Video capture
device
Microprocessor Wireless channel
Host

construct embedded Web server. The architecture of
the AppWeb3.0 is shown in Fig. 2.
Figure 2.Appweb 3.0 architecture.
AppWeb3.0 is open source Web server,
which completely follows GPL(GNU General Public
License) software license agreement. The software
adopts modular design to realize function of the
groupware. This version improves the performance of
the system, introduces the JavaScript framework at
the server side, and supports Windows CE, Windows
7, Vista, Mac OS X, and Free BSD. When running, it
only needs 800K bytes of memory. And more, it adds
JavaScript and Ejscript framework for self
developing, supports dynamic pages make, CGI, SSL
that can be loaded module, ID authentication of the
type of summary, single and multi-threaded
applications, and also supports MVC(model-view-
controller)pattern, providing the ORM(Object
Relational Mapping). So it can be used in the
embedded system perfectly. When carrying on the
AppWeb transplant on the development board, some
environment variables and configuration options need
to be modified.
 We set CC, AR, LD, RANLIB, STRIP
(environmental variables) to the corresponding
location of crosscompilation tool, and set
CC_FOR_BUILD to the location of host. For
example: as at the shellcommand in Linux
system, we input: export
CC=/usr/local/arm/3.4.1/bin/arm-linux gcc,
other similar.
 After setting, we carry on configure, for which
many parameters are provided in the AppWeb
website. The type of the goal board is set as
―arm-s3c2410- linux‖ through ―—host‖
parameter. In order to carry on the cross-
compiling on the PC installed Linux, ―—build‖
parameter set as ―i686-pc-linux‖.
 After completing configuration, we carry out
the commands of ―make‖ and ―make install‖,
then download the homepage required by
AppWeb, the CGI procedure and execution
document of AppWeb to the development
board. Finally, we configure IP and start
AppWeb to visit the server through the IE
browser.
3.2.2. Module for Video Capture based on V4L2
Compared withV4L, V4L2(Video4Linux2)
provides with better extension and flexibility, and
supports more hardware device. However, V4L2 has
made the thorough transformation based V4L, so
both are incompatible to each other. V4L2 provides a
set of unified API for the video application
procedure, which can operate all kinds of different
video capture devices by calling the standard system
function, greatly simplified the development and
maintenance of video system. V4L2 provides the
following interfaces:
• Video capture interface (/dev/video)
• Video overlay interface (/dev/video)
• Video input interface (/dev/video)
• Video output interface (/dev/video)
• Codec interface (/dev/video)
• Effect devices interface (/dev/video)
• Raw vertical blancking interval (VBI) interface
(/dev/vbi)
• Sliced VBI data interface (/dev/vbi)
• Teletext interface (/dev/vtx)
• Radio interface (/dev/radio)
• Radio data system (RDS) interface (/dev/radio)
[5]
In this article, we will cover usage of the
video capture interface, provided by the /dev/video
character device. The flow chart of Video capture is
shown in Figure. 3.
Figure 3.Flow chart of Video capture.

Most of functions of V4L2 is completed
through the ioctl system call, the format of the
grammar is: ioctl (int fd, int request, void *argp).
Here, ―fd‖ is file descriptor, which obtains through
open( ) function; ―request‖ is a command flag, telling
system to do something; ―argp‖ is user data pointer,
transmitting the parameters or receiving the data. The
main functions in the video capture procedure
represent as follows:
• Camera_open( )：call open( ) function to open a
device;
• Camera_get_capability ( ): Through calling ioctl(
) function, we obtain the related information of
the device document using the
VIDIOC_QUERYCAP command flag, and save
to the struct variable of ―struct v4l2_capability‖.
• Camera_cap_init ( ): call ioctl( ) function to set
the video format and frame format by the
command flags of the VIDIOC_S_STD and
VIDIOC_S_FMT in this function. Then, apply
for the frame buffer for the video data, the
number of which is generally not more than 5.
After that, we map the frame buffer to the user
space and add it in queue by the function of
mmap( ), preparing for saving video data.
• Camera_cap_image( ) ： call ioctl( ) function to
caputure vedio data by the command flag of
VIDIOC_STREAM, and pop the frame buffer
whichhas video data from the queue through the
command flag of VIDIOC_DQBUF, then re-add
the frame buffer that has popped data to the
queue tail with the command flag of
VIDIOC_QBUF, so as to collect circularly.
• Camera_close( ) ： use the command flag of
VIDIOC_STREAMOFF to stop video capture,
then call close(fd) fuction to close video device.
3.2.3 Module of Compression and Coding for
Video
In order to transmit the video data via
internet , we have to compress and code the primitive
image before transmission. We use MPEG-4 video
codec here, because the MPEG-4 compression and
coding technology has been already mature and
especically suits for video transmission in the
condition of low bandwidth. When capturing video,
the object in motion is clear and the background is
wheelchair-bed in this system. Since this system has
these characteristics, in order to improve the
compression ratio of video code to achieve a better
performance, we can optimize the algorithm based on
the block match movement estimation in MPEG-4.
The ideals and methods are as follows：
• Identify the movement regions and non-movement
areas. Modifying the MPEG-4 encoder slightly in the
aspect of the movement estimation, we can detect the
movement regions and non-movement areas. The
method is that the macro blocks whose motion vector
is zero are considered as a nonmovement regions and
the rest as moving regions. Certainly, when sampling
two frames as motion detection, the two frames may
not be continual, but we 'd better to take the two
primitive frames. If you take the restructuring frames,
it will has a great distortion with the real scene, and a
bad effect.
• Enhance speed of the motion detection using the
method of sub-regional and recurrent motion
detection. First, the rectangular video frame is
divided into four rectangular areas in the same size,
then carry on motion detection for each region.
According to the necessity of the detection, each
region may also use the same method to carry on the
division and the motion detection again, and the
depth of division is predetermined. If there is a
region whose movement vectors of all the macro
blocks have been determined to be zero, we consider
it as the non-movement region, and no longer do the
division. In the process of doing division of
rectangular region, we must ensure that height and
width are multiples of 16, otherwise, for filling .
• Sub-regional treatment. The MPEG-4 encoder that
codes for each frame is based upon the macro block,
so we can deal with macro block of movement region
and non movement region separately. Regarding the
macro block of non-movement region, we only code
once for it during a period of time, and save it in the
decoding site after decoding. When decoding, we call
the former result of decoding directly. Certainly, in
order to obtain a better effect, it is necessary to
recode for the nonmovement region for a while.
3.2.4 Module of Network Transmission based on
JRTPLIB
JRTPLIB is an object-oriented RTP library,
follows the design of RFC 1889 completely, and
supports many kinds of operating systems, such as
Windows, Linux, FreeBSD, Solaris, Unix and
VxWorks, etc. It adopts the socket mechanism and
solves the network transmission of video streams
well.
3.2.4.1 Data Sending
Figure.4 shows the flow chart of the sending
data.
Firstly, we get the IP address and port
number form the receiving site, then create a session
instance with the class of RTP Session and call the
functions of Create( ) and Set Time stamp Unit( ) to
initialize the session instance. After initializing
successfully, we set the goal address to send by
calling the methods of RTP Session class, which are
the functions of Add Destination ( ), Delete
Destination ( ) and the Clear Destinations ( ). Then
we set the default parameter of the RTP Session

instance by calling the functions of
SetDefaultPayloadType( ), SetDefaultMark( ) and the
SetDefaultTimeStampIncrement( ). At last, the
stream data was sent by calling the method of
SendPacket( ) in the class of RTPSession.
Figure 4.Flow chart of the sending data
Firstly, we get the IP address and port
number form the receiving site, then create a session
instance with the class of RTPSession and call the
functions of Create( ) and SetTimestampUnit( ) to
initialize the session instance. After initializing
successfully, we set the goal address to send by
calling the methods of RTPSession class,which are
the functions of AddDestination( ),
DeleteDestination( ) and the ClearDestinations( ).
Then we set the default parameter of the RTPSession
instance by calling the functions of
SetDefaultPayloadType( ), SetDefaultMark( ) and the
SetDefaultTimeStampIncrement( ). At last, the
stream data was sent by calling the method of
SendPacket( ) in the class of RTPSession.
3.2.4.2 Data Receiving
Figure.5 shows the flow chart of data receiving site.
Figure 5.Flow chart of receiving data
Firstly, we obtain the port number which is
set by users and create an object of RTPSession class;
then we set the receiving mode by calling the
methods of SetReceiveMode( ) and the
AddToAcceptList( ). Secondly, we call the method of
PollData( ) to receive the data packet of RTP or
RTCP that has been sent.Then we can either call the
methods of GotoFirstSource( ) and GotoNextSource(
) to search all sources, or call the methods of
GotoFirstSourceWithDat( ) and
GotoNextSourceWithData( ) to search those source
that carry the data, when searching the RTP source.
The reason is that it is allowed many participants
(source) to exist in the same RTP session. After
detecting the valid data source, we can call the
method of GetNextPacket( ) to get a RTP data
packet, then release it[4]
IV. Voice recognition
There are two essential characteristics for
any effective ASR system: accuracy and speed. In
addition, to meeting these demands, ASR systems
face a number of additional challenges including the
large variance that exists among individual human
speech patterns (e.g. pitch, rate, inflection). A
successful ASR system requires extraordinary
flexibility to accommodate these variances. The
process of ASR typically follows these steps:
1. Acoustic speech is captured by a microphone and
undergoes analog-to-digital conversion
2. Short time frames of speech are converted into
signal features that can be used to discern speech
components
3. Series of frames are mapped to phonemes, the
smallest building blocks of speech, and series of
phonemes are mapped to words.
4. Contextual information is simultaneously
processed to increase the accuracy of the ASR
system. For example, contextual information
would be able to assist in the selection of a series
of homonyms (e.g. merry, marry, and Mary).
The series of words are determined based on the
a probabilistic measure and provided in a format
suitable for the next stage in the application (e.g.
text format).
Sphinx-4 is a state-of-the-art speech
recognition system written entirely in the
Java programming language. It was created via a
joint collaboration between the Sphinx group at
Carnegie Mellon University, Sun Microsystems
Laboratories, Mitsubishi Electric Research Labs
(MERL), and Hewlett Packard (HP), with
contributions from the University of California at
Santa Cruz (UCSC) and the Massachusetts Institute
of Technology (MIT). Sphinx-4 started out as a port
of Sphinx-3 to the Java programming language, but
evolved into a recognizer designed to be much more

flexible than Sphinx-3, thus becoming an excellent
platform for speech research. Live mode and batch
mode speech recognizers, capable of recognizing
discrete and continuous speech.
Sphinx-4 is a very flexible system capable
of performing many different types of recognition
tasks. As such, it is difficult to characterize the
performance and accuracy of Sphinx-4 with just a
few simple numbers such as speed and accuracy.
Instead, we regularly run regression tests on Sphinx-4
to determine how it performs under a variety of tasks.
These tasks and their latest results are as follows
(each task is progressively more difficult than the
previous task):
 Isolated Digits (TI46): Runs Sphinx-4 with pre-
recorded test data to gather performance metrics
for recognizing just one word at a time. The
vocabulary is merely the spoken digits from 0
through 9, with a single utterance containing just
one digit.
(TI46 refers to the "NIST CD-ROM Version of
the Texas Instruments-developed 46-Word
Speaker-Dependent Isolated Word Speech
Database".)
 Connected Digits (TIDIGITS): Extends the
Isolated Digits test to recognize more than one
word at a time (i.e., continuous speech). The
vocabulary is merely the spoken digits from 0
through 9, with a single utterance containing a
sequence of digits.
(TIDIGITS refers to the "NIST CD-ROM
Version of the Texas Instruments-developed
Studio Quality Speaker-Independent Connected-
Digit Corpus".)
 Small Vocabulary (AN4): Extends the
vocabulary to approximately 100 words, with
input data ranging from speaking words as well
as spelling words out letter by letter.
 Medium Vocabulary (RM1): Extends the
vocabulary to approximately 1,000 words.
 Medium Vocabulary (WSJ5K): Extends the
 Medium Vocabulary (WSJ20K): Extends the
 Large Vocabulary (HUB4): Extends the
vocabulary to approximately 64,000 words
Sphinx-4 is an HMM-based speech
recognizer. HMM stands for Hidden Markov Models,
which is a type of statistical model. In HMM-based
speech recognizers, each unit of sound (usually called
a phoneme) is represented by a statistical model that
represents the distribution of all the evidence (data)
for that phoneme. This is called the acoustic
model for that phoneme. When creating an acoustic
model, the speech signals are first transformed into a
sequence of vectors that represent certain
characteristics of the signal, and the parameters of the
acoustic model are then estimated using these vectors
(usually called features). This process is called
training the acoustic models.
During speech recognition, features are
derived from the incoming speech (we will use
"speech" to mean the same thing as "audio") in the
same way as in the training process. The component
of the recognizer that generates these features is
called the front end. These live features are scored
against the acoustic model. The score obtained
indicates how likely that a particular set of features
(extracted from live audio) belongs to the phoneme
of the corresponding acoustic model.
The process of speech recognition is to find
the best possible sequence of words (or units) that
will fit the given input speech. It is a search problem,
and in the case of HMM-based recognizers, a graph
search problem. The graph represents all possible
sequences of phonemes in the entire language of the
task under consideration. The graph is typically
composed of the HMMs of sound units concatenated
in a guided manner, as specified by the grammar of
the task. As an example, lets look at a simple search
graph that decodes the words "one" and "two". It is
composed of the HMMs of the sounds units of the
words "one" and "two":
Constructing the above graph requires
knowledge from various sources. It requires
a dictionary, which maps the word "one" to the
phonemes W, AX and N, and the word "two" to T
and OO. It requires the acoustic model to obtain the
HMMs for the phonemes W, AX, N, T and OO. In
Sphinx-4, the task of constructing this search graph is
done by the linguist.
Figure 6 Search Graph
Usually, the search graph also has
information about how likely certain words will
occur is as shown in Figure 6. This information is
supplied by the language model. Suppose that, in our
example, the probability of someone saying "one"
(e.g., 0.8) is much higher than saying "two"
(0.2).Then, in the above graph, the probability of the
transition between the entry node and the first node

of the HMM for W will be 0.8, while the probability
of the transition between the entry node and the first
node of the HMM for T will be 0.2. The path to "one"
will consequently have a higher score.
Once this graph is constructed, the sequence of
parameterized speech signals (i.e., the features) is
matched against different paths through the graph to
find the best fit. The best fit is usually the least cost
or highest scoring path, depending on the
implementation. In Sphinx-4, the task of searching
through the graph for the best path is done by
the search manager.
As you can see from the above graph, a lot
of the nodes have self transitions. This can lead to a
very large number of possible paths through the
graph. As a result, finding the best possible path can
take a very long time. The purpose of the pruner is to
reduce the number of possible paths during the
search, using heuristics like pruning away the lowest
scoring paths.
As we described earlier, the input speech
signal is transformed into a sequence of feature
vectors. After the last feature vector is decoded, we
look at all the paths that have reached the final exit
node (the red node). The path with the highest score
is the best fit, and a result taking all the words of that
path is returned.
4.1 Sphinx-4 Architecture and Main Components:
When the recognizer starts up, it constructs
the front end (which generates features from speech),
the decoder, and the linguist (which generates the
search graph) according to the configuration
specified by the user. These components will in turn
construct their own subcomponents. For example, the
linguist will construct the acoustic model, the
dictionary, and the language model. It will use the
knowledge from these three components to construct
a search graph that is appropriate for the task. The
decoder will construct the search manager, which in
turn constructs the scorer, the pruner, and the active
list which is shown in Figure 7.
Most of these components represents
interfaces. The search manager, linguist, acoustic
model, dictionary, language model, active list, scorer,
pruner, and search graph are all Java interfaces. There
can be different implementations of these interfaces.
For example, there are two different implementations
of the search manager. Then, how does the system
know which implementation to use? It is specified by
the user via the configuration file, an XML-based file
that is loaded by the configuration manager. In this
configuration file, the user can also specify
the properties of the implementations. One example
of a property is the sample rate of the incoming
speech data.
Figure 7 Sphinx-4 Architecture
The active list is a component that requires
explanation. Remember we mentioned that there can
be many possible paths through the search graph.
Sphinx-4 currently implements a token-passing
algorithm. Each time the search arrives at the next
state in the graph, a token is created. A token points
to the previous token, as well as the next state. The
active list keeps track of all the current active paths
through the search graph by storing the last token of
each path. A token has the score of the path at that
particular point in the search. To perform pruning, we
simply prune the tokens in the active list.
When the application asks the recognizer to
perform recognition, the search manager will ask the
scorer to score each token in the active list against the
next feature vector obtained from the front end. This
gives a new score for each of the active paths. The
pruner will then prune the tokens (i.e., active paths)
using certain heuristics. Each surviving paths will
then be expanded to the next states, where a new
token will be created for each next state. The process
repeats itself until no more feature vectors can be
obtained from the front end for scoring. This usually
means that there is no more input speech data. At that
point, we look at all paths that have reached the final
exit state, and return the highest scoring path as the
result to the application [6]
V. Conclusion
In this paper, we present a voice recognition
system for the robot control based on human voice.
The accuracy reaches a level of 95% which is much
higher than the ones provided by commercial systems
in the market. The core acoustic modeling was
achieved by using the HMM method and more words
than the five commands are built in the linguist

database to improve the accuracy by reading the past
recognized words in the buffer and increase or
decrease of the possible occurrence of the output
words. An embedded voice controlling system is
designed in the paper which is based on S3C2410AL
constructed on ARM9. The programs for video
capturing, image compressing and data transmission
are coded.
References
[1]. ad-hoc robot wireless communication by
Zhigang Wang, MengChu Zhou and Nirwan
Ansari
[2]. Wireless Transmission of RS232 Interface
Signal Based on ZigBee by Guoxin Luo
[3]. The Study of Multi–Robot Communication
of Autonomous Soccer Robots Based on C/S
Mode by XiangZhongfan, WangQiang,
WenHaiou
[4]. A novel embedded System of video capture
and transmission applied in remote
monitoring for wheelchair-bed service
robots by He Qingyun, Liu Jizhong, Zhang
Hua
[5]. voice based robot control by Peter X. Liu,
A. D. C. Chan, R. Chen, K. Wang, Y. Zhu
[6]. A HMM Speech Recognition System Based
on FPGA by Sujuan Ke, Yibin Hou,
Zhangqin Huang, Hui Li
Authors
G.SITA ANNAPURNA received her
master degree in Embedded Systems from JNTU
Kakinada University in Gudlavalleru Engineering
College, Gudlavalleru. She is an Asst Professor of
ECE in Sri Vasvai Institute of Engineering and
Technology, Nandamuru.
B.MAMATHA received her her master
degree in Embedded Systems from JNTU Kakinada
University in Pragati Engineering College,
Surampalem.She is an Asst Professor of ECE in Sri
Vasvai Institute of Engineering and Technology,
Nandamuru.
P.ANNAPURNA received her bachelor
degree in ECE from BIT Institute of Engineering &
Technology. She is an Asst Professor of ECE in Sri
Vasvai Institute of Engineering and Technology,
Nandamuru.

Bq4301381388

More Related Content

What's hot

Viewers also liked

Similar to Bq4301381388

Recently uploaded

Bq4301381388