This article presents a low-cost "smart glove" called HandTalk that can recognize basic hand gestures through flex sensors and convert them to speech using off-the-shelf components like a PIC microcontroller. The glove is intended to serve as assistive technology for the deaf by allowing hand signals to be translated to audio speech or text displayed on a computer. It demonstrates how embedded systems can be effective without expensive components.

1. HandTalk: Assistive Technology for the DUMB
(AMERICAN SIGN LANGUAGE)
Sarji, D.K.
Computer
Volume 41, Issue 7, July 2008 Page(s):84 - 86
Digital Object Identifier 10.1109/MC.2008.226
Summary:One of the many areas in which
embedded systems show great promise is assistive
technologies, which address the special needs of
those with impairments. This article presents
HandTalk, a "smart glove" that can recognize basic
hand gestures and convert them into speech using
low-cost, commercial off-the-shelf (COTS)
components. A low-cost, portable gesture-to-speech
glove prototype demonstrates that embedded systems
don't have to be expensive to be effective.
HandTalk: Assistive Technology for the Dumb
Technology has always been of great help to the disabled and given them a
helping hand to allow them to live a normal and healthy life like others. we
have come up with a novel idea of a glove named Handtalk that will convert
hand movements into text and allow the deaf to express themselves better.
The Handtalk glove needs to be worn on the hand by the deaf or mute person
and depending on the variation of movement, the device will convert it

2. intelligently into text and display it on a nearby computer for the other
person to comprehend it easily.
The Handtalk glove senses the movements through the flex sensors pads
which detect the different patterns of motion and the way the finger curls.
The device can sense carefully each resistance and each movement made by
the hand. Currently the device can convert only few words, but depending
on the success of this device few more additional words may be added later
onto this expressive system.
The Gestures can be converted to voice by using a text to voice convertor
softwares in the computer.
The heart of the system will be the PIC Microcontroller which is responsible
for converting the hand movement signals to the corresponding codes which

3. is further converted into text by the computer.

4. Flexible Bend Sensors
The Flex Sensor is a unique component that changes resistance when bent.
An un flexed sensor has a nominal resistance of 10,000 ohms (10 K). As the
flex sensor is bent the resistance gradually increases. When the sensor is
bent at 90 degrees its resistance will range between 30-40 K ohms.
The flex sensor may be bent greater that 360 degrees depending upon the
radius of the curve. Operating temperature is -45F to 125F.

5. The sensor measures 1/4 inch wide, 4 1/2 inches long and only .019 inches
thick!

6. VOICE
AMPLIFIER/SPEA
KER
Block Diagram OUTPUT(20WAT
TS)
OUTPUT(APR9600)
Power Supply(+5v BOARD
dc supply)
Flex
Sensor 1
+
1
2
V
Flex
P Sensor 2
O Signal converter
W PIC Microcontroller
E
with voltage
PIC 16F877
R Flex follower
Sensor 3
S
U
P
P Flex
L Sensor 4
Y

7. VOICE
AMPLIFIER/SPEA
KER
Block Diagram OUTPUT(20WAT
TS)
OUTPUT(APR9600)
Power Supply(+5v BOARD
dc supply)
Flex
Sensor 1
+
1
2
V
Flex
P Sensor 2
O Signal converter
W PIC Microcontroller
E
with voltage
PIC 16F877
R Flex follower
Sensor 3
S
U
P
P Flex
L Sensor 4
Y

8. VOICE
AMPLIFIER/SPEA
KER
Block Diagram OUTPUT(20WAT
TS)
OUTPUT(APR9600)
Power Supply(+5v BOARD
dc supply)
Flex
Sensor 1
+
1
2
V
Flex
P Sensor 2
O Signal converter
W PIC Microcontroller
E
with voltage
PIC 16F877
R Flex follower
Sensor 3
S
U
P
P Flex
L Sensor 4
Y

9. VOICE
AMPLIFIER/SPEA
KER
Block Diagram OUTPUT(20WAT
TS)
OUTPUT(APR9600)
Power Supply(+5v BOARD
dc supply)
Flex
Sensor 1
+
1
2
V
Flex
P Sensor 2
O Signal converter
W PIC Microcontroller
E
with voltage
PIC 16F877
R Flex follower
Sensor 3
S
U
P
P Flex
L Sensor 4
Y