Currency validation system using mobile

Currency
 Currency is the means of purchasing
through trade. Today, currency generally
refers to printed or minted money.
Sometimes only paper bills are thought
of as currency, while other times coins
are included. Currency involves the
exchange of goods or services for cash.

Currency counterfeit
 Counterfeiting of money is one of the
oldest crimes in history. It was a serious
problem during the 19th century when
banks issued their own currency. At the
time of the Civil War, it was estimated
that one-third of all currency in
circulation was counterfeit

Currency counterfeit detectors

Why currency validation
system
Because of the availability of mobile in
every hand
The customer didn’t have to buy any of the
previous devices only he pick his mobile
phone up and take a photo for the
currency and through our application he
can determine whether that currency
real or false.

Currencies
 Our system applied on Egyptian
currencies

Used Technologies:
 Matlab
 Microsoft visual studio
 Windows phone developer tools

Processing steps
 Taking a photo for the suspected currency
via mobile
 Sending the currency image to the server
to be checked
 Preprocessing of the input image to
remove noise and background
 Currency value determinant
 Extracting special features
 Checking the validity of the currency
 Show result and send it back to the mobile

Step1: taking a photo for the
suspected currency via mobile

 Check to see if the camera is available on the
device
 if not available
 Message ===== > The camera not supported
on the device.
 if available
 Use standard camera if available.
 Otherwise, use front-facing camera on the
device.
 Start image capture
 Save picture as JPEG to isolated storage.

Step2: Sending the currency image
to the server to be checked

 In this steps the mobile send the
captured image to the server using
network
 And the mobile waiting for response

 send image to server by network
Connecting to a TCP Socket Server
 create a socket and connect to the server
by using the System.Net.Sockets APIs.
 Send Request to Server for connect by
Host name and Port Number
 Server listen to any request and Accept On
Connection
 Client Convert image to Array of bytes
 Send Array to server

 Waiting for Result from Server
 Server Received Byte Array
 Convert array of bytes to Image
 Run processing Operations on This
Image

Step3: Preprocessing of the input
image to remove noise and
background
 This step concerns with preparing the
input image for the processing
 ž First, resizing the input image to a
predefined general size
 ž Second, removing the image noise if
exist
 finally, removing the currency
 background

Removing noise
 Using blurring filters to remove noise
from the image

Removing background
 Blobcounter class

Step4: Currency value determinant

 ž In this step we determine the value of
the currency
 ž There was 2 methods
1.MSE
2.comparison method using Surf
algorithm

1- MSE
 Create a matlab function that calculate the
mean square error for the spectrum of the
image
 Binding that matlab function with c#
 ž Crop a rectangle from the upper left
corner
 ž Apply matlab function that calculate MSE
in the cropped part
 ž Test the result and get the range
 ž Determine the currency value

MSE limitations
 ž By examining the MSE method we find
an error rate between 20% & 40%
 That rate come from inference between
200L.E & 100L.E rates
 ž So, we apply the second method .

Comparison method
 In this method the system determine the
value of the currency by comparing the
input currency with a small database
that contains a number of different
currencies
and get the similarity rate for each
comparison
 ž We determine the currency value by
getting the highest similarity rate in
comparison process

Step5: Extracting security features

 After determining the value of
currency, we need to extract the security
features for each currency
 Security marks of currencies are those
marks with which we can determine
whether the currency is true or false

Some important security marks in
the
Egyptian currency :

 Shiny intermittent tape

 Magnetic security tape

 Eye of Horus

 A print using an ink which its color is
changeable when tilting the currency

Step6: Checking the validity of the
currency

 In order to check the validity of the
currency we need to compare the
extracted security marks with another
real currency marks that is saved in the
system
 This process is done using two different
techniques:
1. Neural Network
2. Surf algorithm

Neural network
 Simply we use xor neural network but with
changing the inputs of the network to the
numbers which obtained by counting ones
exits in rows and columns of special parts
in the currency like Shiny intermittent tape
this is called projection.
 In the training process we use back
propagation algorithm to learn neural by
using at least one original currency and
one that is not original .

Neural network
 The ideal output of the original currency
is 1 and the ideal output of the not
original one is 0.
 The network stopped when the number
of epochs is larger than 5000 and the
error is less than 0.001.then we test the
network by a new data.
 We compare the output of the network
by a threshold to determine which it is
original or not.

Surf algorithm
 Convert the input image to integral
image
 Applying fast Hessian filter for
generating interest point
 For each interest point generate surf
descriptor
 Comparison between surf descriptors in
each image

Step7: Show result and send it
back to the mobile
 In this step the mobile receive the result
from the server and show a message to
the user

Limitations
 Camera with more than 5 MP efficiency
 The photo of currency must be taken
with black background
 Our system is not applied in old versions
of currency

Team:
 Ahmad Mohammed Akl
 Abdurrahman Mohammed
 Amira Raft Ahmed Elhmamsy
 ž Ibrahim El-Said Mohammed El-Said
Harhera
 Mai Magdy Mohammad ElKobrosly
 Nada Said El-Said Essa
 ž Nada Mohammed Mohammed Yousef

Currency validation system using mobile

More Related Content

Similar to Currency validation system using mobile (20)

Recently uploaded (20)

Currency validation system using mobile