speech recognition for robotic vehicle control using multi regional language

50
Speech recognition for robotic vehicle control using Multi Regional Language. ABSTRACT The term "robot" for the most part implies some (human-like) appearance exploration instituted some examination issues for creating humanoid robot and one of the noteworthy examination issues is to create machine that have human-like discernment. How is human observation? - The five established human sensors - vision, hearing, touch, smell and taste; by which they percept the encompassing scene. The fundamental objective of our undertaking is to present "hearing" sensor furthermore the discourse union to the Mobile robot such that it is competent to connect with human through Spoken Natural Language (NL). Discourse acknowledgment (SR) is a noticeable innovation, which helps us to present "hearing" and Natural Language (NL) interface through Speech for the Human-Robot communication. So the guarantee of human robot is beginning to wind up a reality. We have picked Mobile Robot, since this kind of robot is getting prominent as an administration robot in the social setting, where the fundamental test is to interface with human.

Upload: tj-ideaforinnovation

Post on 21-Mar-2017

65 views

Category:

Technology


15 download

TRANSCRIPT

Page 1: Speech recognition for robotic vehicle control using Multi Regional Language

Speech recognition for robotic vehicle control using Multi Regional Language.

ABSTRACT

The term "robot" for the most part implies some (human-like) appearance exploration

instituted some examination issues for creating humanoid robot and one of the noteworthy

examination issues is to create machine that have human-like discernment. How is human

observation? - The five established human sensors - vision, hearing, touch, smell and taste; by

which they percept the encompassing scene. The fundamental objective of our undertaking is to

present "hearing" sensor furthermore the discourse union to the Mobile robot such that it is

competent to connect with human through Spoken Natural Language (NL).

Discourse acknowledgment (SR) is a noticeable innovation, which helps us to present

"hearing" and Natural Language (NL) interface through Speech for the Human-Robot

communication. So the guarantee of human robot is beginning to wind up a reality. We have

picked Mobile Robot, since this kind of robot is getting prominent as an administration robot in

the social setting, where the fundamental test is to interface with human.

Two sort of methodology have been decided for Voice User Interface (VUI) execution -

utilizing a Hardware SR framework and another, utilizing a Software SR framework. We have

taken after Hybrid engineering for the general apply autonomy outline and correspondence with

the SR framework; likewise made the punctuation for the discourse, which is decided for the

automated exercises in his field. The configuration and both usage methodologies are exhibited

in this report.

One of the vital objectives of our task is to present appropriate UI for tenderfoot client and our

test arrangement is planned by our undertaking objectives; so we have likewise directed an ease

of use assessment of our framework through learner clients. We have performed tests with

straightforward and complex sentences for various sorts of apply autonomy exercises;

Page 2: Speech recognition for robotic vehicle control using Multi Regional Language

furthermore broke down the test result to discover the issues and confinements. This report

exhibits all the test outcomes and the discoveries, which we have accomplished all through the

venture.

CHAPTER 1

INTRODUCTIONThis project reports a couple of outcome of progressing research extend that means to investigate

approaches to command a modern robot utilizing the human voice with territorial dialect. These

elements can enthusiasm with a few modern, research facility and clean‐room applications,

where a nearby participation among robots and people is alluring. Apply autonomy Is the branch

of mechanical building, electrical designing and software engineering that arrangements with the

configuration, development, operation, and use of robots, and also PC frameworks for their

control, tactile criticism, and data preparing. These advances manage robotized machines that

can replace people in unsafe situations or assembling forms, or take after people in appearance,

conduct, and/or insight. A hefty portion of today's robots are motivated by nature adding to the

field of apply autonomy.

The idea of making machines that can work self-sufficiently goes back to traditional times,

however scrutinize into the usefulness and potential employments of robots did not become

generously until the twentieth century. All through history, apply autonomy has been regularly

seen to copy human conduct, and frequently oversee assignments in a comparative design.

Today, mechanical autonomy is a quickly developing field, as innovative advances keep;

inquiring about, outlining, and building new robots fill different functional needs, whether

locally, monetarily, or militarily. Numerous robots do tasks that are unsafe to individuals, for

example, defusing bombs, mines and investigating wrecks.

There are numerous sorts of robots; they are utilized as a part of various situations and for a

wide range of employments, in spite of the fact that being extremely differing in application and

structure they all offer three essential similitude’s with regards to their development Robots all

have some sort of mechanical development, an edge, frame or shape intended to accomplish a

Page 3: Speech recognition for robotic vehicle control using Multi Regional Language

specific undertaking. For instance, a robot intended to traverse substantial earth or mud, may

utilize caterpillar tracks.

1.1 Problem Formulation

Since as we seen numerous customary robotic frameworks .As automated framework comes

generally with programmed operations as we seen such frameworks like line devotee robot,

modern assembling robots and so on, coming to controlled automated frameworks,

1.2 Existing System

The best in class in tangible insight for robots will need to advance through a few requests of

greatness in the event that we need the robots working in our homes to go past vacuum-cleaning

the floors. On the off chance that robots are to work adequately in homes and other non-modern

situations, the way they are told to play out their employments and particularly how they will be

advised to stop will be of basic significance. The general population who associate with them

may have practically zero preparing in apply autonomy, thus any interface should be to a great

degree natural. Sci-fi creators likewise commonly accept that robots will in the end be equipped

for speaking with people through discourse, motions, and outward appearances, as opposed to a

summon line interface. In spite of the fact that discourse would be the most characteristic path

for the human to impart, it is unnatural for the robot. It will likely be quite a while before robots

cooperate as normally as the anecdotal C-3PO.

1.3 Proposed System

This project reports a couple aftereffects of a progressing research extend that expects to

investigate approaches to charge a mechanical robot utilizing the human voice with provincial

dialect. This component can enthusiasm with a few mechanical, research center and clean‐room

applications, where a nearby participation among robots and people is alluring.

1.4 Methodology

Page 4: Speech recognition for robotic vehicle control using Multi Regional Language

A showing is introduced utilizing two mechanical robots and a (PC) furnished with a sound

board and a headset mouthpiece. The exhibition was coded utilizing the Microsoft Visual Basic

and C#.NET 2010 and connected with two straightforward robot applications: one equipped for

picking‐and‐placing objects and going to predefined positions, and the other fit for playing out a

basic direct weld on a work‐piece. The discourse acknowledgment linguistic use is determined

utilizing the sentence structure developer from the Microsoft Speech SDK 5.1. This anticipate

likewise acquaints the ideas of discourse with content interpretation and voice acknowledgment,

and shows how these components can be utilized with applications fabricated utilizing the

Microsoft.NET structure.

Page 5: Speech recognition for robotic vehicle control using Multi Regional Language

CHAPTER 2

LITERATURE SURVEY

1. It is conceivable to appraise the commotion by utilizing data on the robot's own particular

movements and stances, in light of the fact that a sort of movement and signal creates just about

the same example of clamor without fail. In this paper, it is portrayed as voice acknowledgment

control framework for robot (VRCS) framework which can vigorously perceive voice by grown-

ups and youngsters in uproarious situations. . It is assessed that the VRCS framework in a

correspondence robot put in a genuine loud environment. Voice is caught utilizing a remote

receiver. To stifle obstruction and commotion and to weaken resonation, this paper actualized a

multichannel framework comprising of an exception vigorous summed up side-projection

canceller strategy and a component space clamor concealment utilizing MMSE criteria. Voice

movement periods are identified utilizing GMM-based end-point recognition [1].

2. Individual automated assistants help decreasing the manual endeavors being put by people in

their everyday errands. This paper executes a voice-controlled individual colleague robot. The

human voice charges are given to the automated aide remotely, by utilizing a savvy cell

telephone. The robot can perform distinctive developments, turns, begin/stop operations and

move an item starting with one place then onto the next. The voice orders are handled

continuously, utilizing an online cloud server. The discourse signal charges changed over to

content structure are conveyed to the robot over a Bluetooth system. The individual associate

robot is created on a small scale controller based stage and can know about its present area. [2].

3. The Intelligent service robot in indoor environment with voice acknowledgment and tele-

presence capacities accepts that this model of service robot is utilized to convey a drinking

container to the perceived client from the beginning stage to the destination. The advancement

Page 6: Speech recognition for robotic vehicle control using Multi Regional Language

and acquaintance of 3 fundamental modules with keep away from deterrent, to perceive charges

and tele-presence capacities in light of Google App Engine. The discovery of snags, which is in

view of the data from the 3 ultrasonic sensors, utilizes an Interval sort 2 Fuzzy Logic for

receiving instability. The fundamental commitment of this examination is to give a complete

system of correspondence model between primary controllers to different modules in

administration robot [3].

4. The upsides of discourse enacted robots are sans hands and quick information info operations.

In future, it is normal that discourse acknowledgment frameworks will be utilized as man-

machine interface for robots in restoration, diversion and so forth. The discourse

acknowledgment framework is prepared in a manner that it perceives characterized summons

and the composed robot explores taking into account the direction through the Speech

Commands. The medium of connection amongst people and PCs is on the preparing of discourse

[4].

5. The impact of haptic input quality on tele-operated undertaking execution and control exertion

has been evaluated for a run of the mill (dis-)get together assignment in a hard-to-hard

environment, surely understood to the administrator. Four basic subtasks have been recognized,

in particular: 1) free-space movement,2) contact move, 3) compelled translational, and 4) obliged

rotational errands. [5].

6. Voice-commands are given as a contribution to the Android OS based stage which are

handled and transmitted to an online cloud server progressively. The voice-commands changed

over to content structure are then conveyed to an installed Bluetooth module of the robot,

utilizing the Blue-tooth module of the advanced cell. The Bluetooth module of the robot gets the

commands as content and after that sends these to the miniaturized scale controller board for

further preparing. The pace control system controlled through a product guarantees halting the

robot before a pre-set differential deferral time [6].

7. The robot depends on the doctor's facility WLAN for remote control and transmission of

information in both headings between the robot and the control room. Lifting, taking care of and

transporting radioactive materials inside the healing center is one of the essential capacities did

by the robot. This capacity is critical to decrease the presentation of doctor's facility staff to

Page 7: Speech recognition for robotic vehicle control using Multi Regional Language

hurtful radiation. Another capacity of this robot is varying media tolerant specialist

correspondence notwithstanding when the specialist can't be physically at the clinic. If there

should arise an occurrence of WLAN disappointment, the robot can be controlled physically

utilizing a joystick [7].

Chapter 3

HARDWARE AND SOFTWARE REQUIREMENTS

In the wake of experiencing the venture detail we have touched base at the accompanying

Hardware and Software necessities. This part gives insights about different hardware

components and software utilized as a part of execution of c# based multi language voice

empowered robot.

3.1 Hardware Requirements

c# based multi language voice empowered robot requires the following hardware components:

ATMEGA16 microcontroller

Motor Driver L293D

CC2500 x-bee

12 volt 1.2Ah battery

Dc motor

The accompanying areas depict these hardware components and clarify why we are utilizing

these specific components as a part of our c# based multi language voice empowered robot.

3.1.1Atmega16 Microcontroller

Atmega-16 also belongs to same family. It also performs the same function of ATmega-8.

ATmega16 is an 8-bit-microcontroller. Atmega16-is mainly depends RISC and CISC

Architecture which includes 135 set of instructions. Oscillation frequency is of 16MHz.

Page 8: Speech recognition for robotic vehicle control using Multi Regional Language

 

fig 3.1 atmega 16

Pin diagram of Atmega-16

Fig 3.2 Pin configuration of atmega-16

Pin no 1 and 2 is connected to timer 0 and 1. Pin no 3 and 4 is connected to positive

and negative inputs.

Pin no 5 to 8 are connected to Serial-Peripheral-Interface. Pin no 9, 10 and 11 are

connected to reset, VCC and to gnd. Pin no 18 and 19is connected to PWM channels

Pin no 22 and 23 are connected to interface of TWI and 24 to 27 are connected to

interface of JTAG

Page 9: Speech recognition for robotic vehicle control using Multi Regional Language

ADC channels are connected from Pin no 33 to 40.

Features:

Flash memory is of 32KB

1KB of static Ram

512-bytes of EEPROM

EEPROM flash memory is 10,000-100,000, individually.

GPIO:

The general purpose input and output pin is a multiplexed pin which can either send or receive

the information or can be configured as uart interrupt pwm or analog channel the mainly there

are four ports in atmega-8 port a,b,c,d to do general input or output operation specified by port

direction setting register or special purpose register can be enabled to use uart ,pwm, adc etc

when enabling this disables the general input and output operation .fig3.2 shows the specific

registers used for a single port.

Fig.3.3 Different port operation

If port direction setting register is set to logic high the corresponding port is configured as output

if it is set to logic low then it’s configured as input in 1 byte mode pinx is used to accesses the

single pin in the corresponding port and whatever the data need to be read or written must be sent

through portx.

UART:

In most of the embedded microcontroller uart is used to convert the parallel data into

serial and serial into parallel to be sent one by one for the long range most of the controller

supports the multi processor communication and three modes of communication full duplex

mode can only be done with the help of inbuilt serial interrupt two communication can be done

Page 10: Speech recognition for robotic vehicle control using Multi Regional Language

one is synchronous mode another is asynchronous this mode is known as independent mode

clock generated by internal baud rate generator gives the required clock for communication

general format of communication consist of one start bit usually not used one stop bit in-between

eight bit data and a parity bit to check the error during transmission or reception of packet.fig 3.3

shows the time diagram of uart for communication.

Fig.3.4 Asynchronous data transfer timing diagram

3.1.2 Voltage regulatorTo give the correct voltage to the circui not intend to damage the controller or any other digital or

analog parts ists mandotry to provide the required power no more or less than the anticapated to achive

this a set of five volt regulators are used which can power up the corresponding device as the motor runs

form 9 volt to 12volt there is no need of reducing the incoming voltage to do with this 7805 constant

5volt 1 amp load driving capability and input range from 6 to 36 voltage is used it has three terminal pin

one is input pin two is common pin three is output from where constant 5 volt is supplied to the devices in

order to keep the constant supply without thermal run away additional heat desipater must be attached to

the device

Page 11: Speech recognition for robotic vehicle control using Multi Regional Language

Figure 3.5 Voltage Regulators (7805)

Usually monolithic voltage regulator is available as 3 terminals IC7805 as shown in below figure. The 3

terminals are denoted as IN (input), COM (common), OUT (output). This +5V regulator is useful in

power up to 500mw.It must have a heat sink for high current. A 1mf high quality and tantalum capacitor

should be placed from output to ground for stability. By using this regulator circuit we are deriving 5v

from 12v battery.

3.1.3 Gear Dc Motor

At whatever point a mechanical autonomy specialist discuss making a robot, the primary thing

rings a bell is making the robot precede onward the ground. What's more, there are constantly

two choices before the architect need to utilize a DC or stepper motor. With regards to speed,

weight, size, cost. DC motors are constantly favored over

stepper motor. There are numerous things which can

be done when dc motors are connected with

microcontroller. For instance rate of motor can be

controlled and control the heading of revolution, can likewise do encoding of the turn made by

DC motor i.e. monitoring what numbers revolution are made by motor .

Fig: 3.6 internal architecture of dc motor

Page 12: Speech recognition for robotic vehicle control using Multi Regional Language

As dc motor don’t have much torque to drive a payload with desired rpm to avoid this problem

sets of gear are used to drive the motor which consist of driver gear idle linking gear and driven

gear which will be connected to shaft of the motor then it can be connected to wheel to drive the

pay load.

Fig. 3.7 gear combination

DC Motor with Gear 

The DC motors don't have enough torque to drive a robot directly by connecting wheels in it. 

Gears are used to increase the torque of dc motor on the expense of its speed. 

Mathematical interpretation: 

Rotational power (Pr) is given by:

Pr =Torque (T) * Rotational Speed ()

 Pr is constant for DC motor for a constant input electrical power. Thus torque (T) is inversely proportional 

Speed  ().  Thus to increase the value of torque we have to loss speed. 

3.1.4 L293D Dual H-Bridge Motor Driver

L293D is a double H-Bridge motor driver this ic can drive two dc motors in both direction with

fix course of movement by interfacing all i/o to microcontroller and up to four motors can be

controlled in fixed direction this driver has yield current of 600ma and up to 1.2A for every line

connected to motors mainly this ic have internal freewheeling diode which protects from back

emf generated during the rotation and works with wide input from

4.5v to 36v.

Page 13: Speech recognition for robotic vehicle control using Multi Regional Language

Fig. 3.8 connection diagram l293d to motor and input function

In the Fig. 3.14 to connect a single motor to driver and to microcontroller o/p1 and o/p2 are

connected to motor i/p pin a,b and en will be connected to microcontroller direction signal will

be given to a and b input en signal must be high additionally vcc1 must be connected to 5volts

vcc2 must be connected to 12 volt and all four ground additional pull of 10k must be connected

to i/o line of microcontroller to ensure the perfect operation of driver.

3.1.5 Battery

Power bank are major source to run the prototype here 12volt 1.2ampere/hour sealed lead acid

battery is used to give constant supply for continuous operation of the system compared to

conventional lead acid battery this bank won’t spill and acid and no need to put distilled water an

additional charger need to charge the battery 14.1 volts is knows as battery full and 10 volts is

known as low battery discharging the battery beyond the level make the battery unusable.

Additional voltage reducer ic 7805 must be used to drive the digital part.

3.1.6 Zig-bee Module

Zig-bee is an wireless transceiver used to send and receive data at maximum communication

speed of 250Kbps the name zig-bee replicates for the bee hive for the mesh network by point to

point communication advantage of using zig-bee over other conventional wireless is it provides

long range communication personal area network support long battery life , beacon option for

Page 14: Speech recognition for robotic vehicle control using Multi Regional Language

remote locating usually zig-bee is used for data acquisition and transmission zig-bee operates at

3.3v and low stand by idle current communication frequency is of standard permissible for

wireless test purpose at2.4Ghz zig-bee can communicate with one device at a time or multiple

devices depending on pan here zig-bee is used to send data from pc to robot by using serial

communication operating at nine thousand six hundred baud rate eight bit data, one stop bit, not

start no parity.

3.1.7 Rs 232 Cable and Max 232 Ic

Most of the microcontroller and microprocessor are cmos and mainly operates at logic level five

volt and zero as logic high and logic low if this system wants to communicate with other ttl or

cmos device of different logic level there may be chance of failure in communication due to

different logic levels and may be chance of getting damaged due to high voltage level hence the

rs232 standard is been used but the major problem once again is logic levels for this reason a ic

maxima 232 is used which consist of voltage inverting and doubling done with te help of 4 o.1

microfarad ceramic disc capacitor connected to pin1 and 3 pin2 and vcc pin 4 and 5 and pin 6

and ground each max 232 can communicate with the two devices of different logic level usually

db-9 connector is used to do the physical connection between sender and receiver db-9 consist of

9 pins upper row consist 5 pins in which pin2 is either sender or receiver depending on type of

connector pin3 is either sender or receiver from there 5th pin is ground which is a mandatory

connection to be done ,for successful connection rest of the pins are known as handshaking

signal for multiprocessor communication.

Max 232 IC Is Utilized to coordinate the working voltage between Zigbee module and PC for

serial communication.

3.1.8 Internal Structure of Max232

MAX232 requires four capacitors ranging from 1 to 10μF. The most widely used value for these

capacitors is 10μF.

Page 15: Speech recognition for robotic vehicle control using Multi Regional Language

Fig: 3.9 Pin diagram of MAX232

3.2 SYSTEM SOFTWARE DESIGNThe software requirements for designing of multi-language robot is as

follows Microsoft visual studio 2010.net

AVR studio version5

SINA programmer

3.2.1 MICROSOFT VISUAL STUDIO 2010.NET:

C# is a dialect of Microsoft visual studio that allows the developers to create specific application

which includes both graphical and code end which can be connected to database.

C# runs only on windows machine by using C# developers can create windows based client

application, data base integration Xml for web related services and many more features are

included by using visual studio frame work user can compile, debug and build the application in

this project c# application is used at receiver side were corresponding person accesses to the data

base of the patient to view the recorded log of the status.

3.2.2 .NET SYSTEM STAGE DESIGN

Page 16: Speech recognition for robotic vehicle control using Multi Regional Language

C# will be executed repeatedly on .NET Structure, is vital piece of operating system which

consolidates a backend execution system called the typical thread runtime which have inbuilt

libraries which is accessed by class. An overall mark is the reason which makes execution and

headway circumstances in which code and inbuilt add on participate reliably. Coding done in C#

is aggregated and transitional lingo that changes with the CLI point of interest. The inline code

and sources, for instance, bitmaps and strings, are secured to a plate in an exe record assembled a

get, normally with a development of .exe or .dll. Parties contain and demonstrate that gives

information about the party's sorts, interpretation, society, and security essentials.

Exactly when the C# framework is executed, the social gathering is stacked into the CLR, and

which may take distinctive exercises in perspective of the details in appear. By then, security

necessities must be fulfilled, CLR done at the last possible second (JIT) get-together to change

over the IL code to neighborhood machine rules. The CLR moreover gives distinctive

organizations related to customize waste social occasion, unique case dealing with, and resource

organization. Code that is executed by the CLR is as a less than dependable rule suggested as

"directed code," instead of "unmanaged code" that can be accumulated to neighborhood

machine.

3.2.3 AVR Studio Version5:

Atmel® AVR Studio® 5 is the integrated development environment (IDE to develop the

firmware for embedded microcontroller of Atmel family by using this tool user can write

compile and debug the code this tool incorporates with multiple features like watch window to

view the registers status peripherals’ widows shows the internal operation of hardware in the

microcontroller.

3.2.4 Steps Involved Generating Project Hex File:

Step 1:

Page 17: Speech recognition for robotic vehicle control using Multi Regional Language

Fig 3.10: Creation of a New Project Step 2:

Fig 3.11: Creation of a New Project Step 2:

Page 18: Speech recognition for robotic vehicle control using Multi Regional Language

Fig 3.12: Selection of Target Device for Which the Hex File is to

be generated

Step3:

Page 19: Speech recognition for robotic vehicle control using Multi Regional Language

Fig 3.13: Write Code With C.

Step4: click on run compile, no errors result.

Step5: build target generate hex file.

3.2.5 Sina Programmer

Fig 3.14: Shows Front View of Sina Programmer

Sina Programmer is burner programming with straightforward client additionally joins an AVR

Wire number cruncher. It utilizes enhanced AVRDUDE 5.10 and underpins new gadgets and

new records just the accessible ports and baud rate determination is conceivable. It is a

standalone straightforward programming and no establishment is select the gadget sort as

ATmega16 and set the Software engineer to USB asp and port to USB (as seen at the base of

window).Click on scan under Hex Document and find the .Hex record to snap System catch

under bits can likewise be customized.

Page 20: Speech recognition for robotic vehicle control using Multi Regional Language

Chapter 4

Methodology

C# based multi language voice controlled robot consists of transmitter unit and receiver

unit. They are connected together wirelessly. This chapter describes the working principle and

flowchart of the system.

4.1 c# and processing This block gives the in detailed function of speech processing on C# using visual studio

platform speech input is given though condenser microphone this input is converted into

electrical signal and given to sampling process after this converted digital values is given to c#

for processing depends on the trained processes code recognize the give audio sample and

depends on the input output will be sent to wireless unit via RS232 port.

Page 21: Speech recognition for robotic vehicle control using Multi Regional Language

Fig.4.1 speech and touch input to C#

4.2Receiving UnitThis module consist of atmega-16 microcontroller which perform all the controlling and

monitoring operation for the corresponding command received by uart through cc2500 atmega-

16 operating at 16 mhz with flash rom of 16kb and ram of 2kb and course diversion is done

using two 30 rpm geared dc motor is used to give the robot the mobility operation which is

driven by the microcontroller with the help of l293d h-bridge motor.

WINDOWS OPERATING

SYSTEM

SPEECH INPUT CONDENSER

MIC

SPEECH RECOGNIZING

AND PROCESSING

VISUAL STUDIO

CC2500

X-BEE

ATMEGA-16 MICROCONTROLLER

12 V 1.2 AH

SLA BATTERY

CC-2500 X-BEE MODULE

Page 22: Speech recognition for robotic vehicle control using Multi Regional Language

Fig.4.2 receiving unit

Fig 4.3 Description of Various Parts of Robotic Vehicle.

L293D H-BRIDGE MOTOR DRIVER

MOTOR1

MOTOR2

Left wheel attached to

Motor 2

7805 REGULATOR

Right wheel attached to

Motor 1

Free wheel front

Chassis

Page 23: Speech recognition for robotic vehicle control using Multi Regional Language

1. straight 2.straight

Forward (mundhe) for char “A” Backward (hindhe) for char “B”

Fig 4.4 Movement of Robot Forward Fig 4.5 Movement of Robot Backward

3. Sharp Right 4. Sharp left

Right (balha) for char “C” Left (yada) for char “C”

Page 24: Speech recognition for robotic vehicle control using Multi Regional Language

Fig 4.6 Movement of Robot Right Fig 4.7 Movement of Robot Left

5. Stop

Stop (nillu) for char “E”

Fig 4.8 Stop No Movement

From the above fig we can analyze the movement and directions of motor for respective commands.

Page 25: Speech recognition for robotic vehicle control using Multi Regional Language

4.4 Circuit Diagram

In this circuit atmega-16 is used to course the direction of the robot as per the command given by

the client these command are received by cc2500 wireless connected to rxd and txd pin of

microcontroller configured as uart operating at nine thousand six hundred baud rate eight bit data

one stop bit no start no parity any command sent by the client will be received through uart and

then processed.

Fig.4.9 hardware connection diagram

2 12volt 30 rpm geared dc motor is used to move the chair in desired direction driven by l293d

h-bridge motor driver as dc motor is analog mechanical and works at 12 volt consuming 200to

300 ma of current and mainly it generates back emf while in motion if this motor is directly

connected to microcontroller means controller will get damaged because its digital operates at

5volt and can supply maximum of 40 ma source current per input and output so l293d gets the

information from microcontroller and change the course of robot.

Page 26: Speech recognition for robotic vehicle control using Multi Regional Language

4.3 Flow Chart of the system

speeech process with c#

NO

YES

YES

NO

START

RUN SPEECH FRAME WORK

INITIALIZE UART FOR PACKET TRANSFER

RUN SPEECH ENGINE

IF SPEECH INPUT

DETECTED

IF COMMAN

D IS FOWARD

SEND ‘A’ TO ROBOT

A

C

B

PROCESS THE SPEECH

Page 27: Speech recognition for robotic vehicle control using Multi Regional Language

YES

NO

YES

NO

Fig.(a) Flow chart of speech processing

SEND ‘B’ TO ROBOT

IF COMMAND

IS LEFT

SEND ‘D’ TO ROBOT

A

B

C

IF COMMAND

IS BACKWAR

D

IF COMMAND

IS RIGHT

SEND ‘C’ TO ROBOT

IF COMMAND

IS STOP

SEND ‘E’ TO ROBOT

Page 28: Speech recognition for robotic vehicle control using Multi Regional Language

ROBOT CONTROL FLOW

NO

YES

YES

NO

INITIALIZE UART AT 9600 BAUDRATE 8BIT DATA ONE STOP

BIT NO START NO PARITY

IF COMMAN

D RECIVED

IF COMMAN

D IS ‘A’

MOVE ROBOT FORWARD

AC

E

PROCESS THE COMMAND

START

Page 29: Speech recognition for robotic vehicle control using Multi Regional Language

YES

NO

YES

NO

Fig.(C) Flow chart of robot movement control

MOVE ROBOT REVERSE

IF COMMAND

IS ‘C’

MOVE ROBOT RIGHT

A

C

IF COMMAND

IS ‘B’

IF COMMAND

IS ‘D’

MOVE ROBOT LEFT

IF COMMAND

IS ‘E’

STOP ROBOT

EE

Page 30: Speech recognition for robotic vehicle control using Multi Regional Language

Chapter 5

Results and Discussion

With modernization in technologies and multiple research in the field of robotics from domestic

to military usage has gone much forward in this project robot movement is controlled using

voice as command input but the special feature is using multi-language user interactive input

processed on C# platform using visual studio command is sent through 2.4ghz cc2500 wireless

transceiver unit below details give the brief function of robot.

The voice processed for the robot control is as described in below steps

(Module pic)

Fig 5.1: full integrated prototype

Step-1:

This the main screen to configure the com port by giving the desired port no and need to open

the port to start the process after this process user need to turn on the speech recognizer by

ticking it otherwise only touch control through cursor can be done.

Page 31: Speech recognition for robotic vehicle control using Multi Regional Language

Fig 5.3: verbal control framework

Step-2:

In this step user has given the voice command “mundae” means forward in English for this

command a char “A” is sent through the serial communication to wireless unit which will be

transmitted to robot to movement.

Page 32: Speech recognition for robotic vehicle control using Multi Regional Language

Fig 5.4: command for forward movement

Step-3:

In this step user has given the voice command “Hindhae” means backward in English for this

command a char “B” is sent through the serial communication to wireless unit which will be

transmitted to robot to movement.

Fig 5.5: command for backward movement

Step-4:

In this step user has given the voice command “bala” means right in English for this command a

char “C” is sent through the serial communication to wireless unit which will be transmitted to

robot to movement.

Page 33: Speech recognition for robotic vehicle control using Multi Regional Language

Fig 5.3: command for right movement

Step-5:

In this step user has given the voice command “yada” means left in English for this command a

char “D” is sent through the serial communication to wireless unit which will be transmitted to

robot to movement.

Fig 5.3: command for left movement

Page 34: Speech recognition for robotic vehicle control using Multi Regional Language

Step-6: In this step user has given the voice command “nillu” means stop in English for this

command a char “E” is sent through the serial communication to wireless unit which will be

transmitted to robot to stop the movement.

Fig 5.3: command to the stop movement

Result & Analysis

Page 35: Speech recognition for robotic vehicle control using Multi Regional Language

Applications:

It has wider applications, by interface a metal detector .It can be used to sense mine in

war fields.

It can be used to travel through tunnels which are so congested places where people cant

travel.

It can be used for industrial applications to carry heavy metals and industrial materials

with in the industrial environment.

By interfacing a webcam .it can be used to monitoring of enemy movements.

Industrial motor speed controlling.

Home automations.

Robotic applications.

Advantages:

Page 36: Speech recognition for robotic vehicle control using Multi Regional Language

This system is compact and flexible so that it can travel over a dense area.

Robot vehicle can be controlled either by button events or by speech events.

Provide more convenience for user in case of language of communication and can be

controlled from long distance.

It can be controlled over 10-100 meters of distance from the transmitting end.

This project provides a best GUI (graphical user interface) developed with platforms of

visual studio.net 2010 with back end coding with C#.

Chapter 6

Conclusion and Scope for future work

This part gives the last conclusion drawn from the proposed work and the extension for future

work.

6.1 Conclusion This framework pays to the self-reliance. This decreases the manual exertion for an

achieving and recognizing the order for controlling the movement of a robot by determined

charges. Further that, the improvement of this model is finished with less cost and reasonable.

This robot can be utilized to do remote surveillance. We are actualizing programmed robot it has

different preferences. We can likewise include new innovation in this robot.

This system can be made highly efficient and effective. The setup for maintaining the

robot will be a onetime investment for any real life application. The motor drive and control

arrangement of the model astute robot has been exhibited. The microcontroller based voice

worked keen robot will bring more comfort in hazard environment surveillance, also the robot

Page 37: Speech recognition for robotic vehicle control using Multi Regional Language

has function of rotating wheel by some defined angle that will be very useful for the user for

taking left or right turn.

6.2 scope for future work

The range of control could be increased by using long range wireless modules like Wi-Fi,

etc.

Android based control with RTP based video rendering can be done.

REFERENCES

[1] Stephen Cook, Speech Recognition HOWTO

[2] Fatih AYDINOĞLU, Robot Control System with Voice Command Recognition, Yıldız

Technical University, Department of Computer Engineering

[3] P. M. Grant, Speech recognition techniques

[4] Sahar E. Bou-Ghazale, Member, IEEE, and John H. L. Hansen, Senior Member, IEEE, A

Comparative Study of Traditional and Newly Proposed Features for Recognition of Speech

Under Stress

[5] M. Chetouani, B. Gas, J.L. Zarader, C. Chavy, Neural Predictive Coding for Speech

Discriminate Feature Extraction : The DFE-NPC, Laboratories des Instruments et Systems d’Ile

de France

Page 38: Speech recognition for robotic vehicle control using Multi Regional Language

[6] Raymond Low and Roberto Togneri, Speech Recognition Using the Probabilistic Neural

Network, The University of Western Australia, Department of Electrical and Electronic

Engineering

[7] M. M. El Choubassi, H. E. El Khoury, C. E. Jabra Alagha, J. A. Skaf and M. A. AlAlaoui,

Arabic Speech Recognition Using Recurrent Neural Networks, Faculty of Engineering and

Architecture – American University of Beirut