Prepared By: Meera S. Gohil Hansa J. Gajera: V.V.P. Engg. College

Neural network
Prepared by:
Meera S. Gohil
Hansa J. Gajera
V.V.P. ENGG. COLLEGE

Why would any one want
a ‘new’ sort of computer
 GOOD AT  NOT SO GOOD AT
 Fast arithmetic  Interacting with
 Doing precisely what noisy data or data
the programmer from environment
programs them to do  Massive parallelism
 Fault tolerance
 Adaptive to
Analogy with brain
An Introduction to the
Neural Networks
 What is Neural Networks?
An Artificial Neural Network (ANN) is an
information processing paradigm that is inspired by
the way biological nervous systems, such as the
brain, process information.
 Where can neural network systems help?
I. where we can't formulate an algorithmic solution.
II. where we can get lots of examples of the
behavior we require.
III. where we need to pick out the structure from
existing data
Biological –Artificial
Neuron
Design
 Layers
 Communication & Types of connection
Inter Layer connection
Intra Layer connection
 Learning
Learning Laws
 Hubb’s Rule
 Hopfield Rule
 The delta Rule
 Kohonen’s Learning Rule
THE BACK –PROPAGATION NEURAL
NETWORK
(MULTI LAYER PERCEPTRON)
Operation of mathematical model
Sigmoid function
Topology of multiplayer perceptron
Training and memory of neural network
BPN training algorithm
The training algorithm for a BPN consists of the following steps:
 Selection and Preparation of Training Data

 Modification of the neuron connection weights
 Repetition
 Running
 Hazards
• ition
• 1.1 Main Purpose
LeavesRecognition is a neuronal network based java application/applet to recognize images of leaves accordingly to a previously trained
Backpropagation Network.
The intention of this application is to give the user the ability to administrate a hierarchical list of leaf images, where he can perform some sort of edge
detection to identify the individual tokens of every image. This tokens will then be the basis of the neuronal network calculations to make it possible to
recognize a unknown leaf image and specify the species it belongs to.
Leaves recognition
1.1 Main Purpose
LeavesRecognition is a neuronal network based java

application/applet to recognize images of leaves accordingly
to a previously trained Backpropagation Network.
The intention of this application is to give the user the

ability to administrate a hierarchical list of leaf images,
where he can perform some sort of edge detection to identify
the individual tokens of every image. This tokens will then be
the basis of the neuronal network calculations to make it
possible to recognize a unknown leaf image and specify the
species it belongs to.
2. Theoretical Background
2.1 Image edge detection
The Prewitt edge detection filter uses these two 3x3

templates to calculate the gradient value:
-1 0 1 111
-1 0 1 000
-1 0 1 -1-1 -1
X Y
| a1 a2 a3 |
| a4 a5 a6 |
| a7 a8 a9 |
where:
a1 .. a9 - are the grey levels of each pixel in the
filter window
X = -1*a1 + 1*a3 - 1*a4 + 1*a6 - 1*a7 + 1*a9
Y = 1*a1 + 1*a2 + 1*a3 - 1*a7 - 1*a8 - 1*a9
Prewitt gradient = SQRT(X*X + Y*Y)
All pixels are filtered. In order to filter pixels

located near the edge of an image, edge pixels
values are replicated to give sufficient data.
2.2 Thinning
The used thinning algorithm here

processed the image recursivly and
minimizes the found lines to a one-pixel
wide one by comparing the actual pixel
situation with specific patterns and then
minimizes it.
2.3 Leaf Image Token
Green line: The shape of the
leaf image after a successful
edge detection & thinning.
Red Square: This square

represents a point on the
shape of the leaf image from
which we are going to draw a
line to the next square.
Blue line: The compound of

the center of two squares
from which we are going to
calculate the cosinus and
sinus angle. Such a blue line
is a representation of a leaf
token.
On the left hand side you see
a small image of the right-
angled triangle which
represents a token of a single
leaf image. Here it should be
clear now that the angles A
and B are the two necessary
parts which will be fit into the
neuronal network layers.
With this two angles we can

exactly represent the direction
of the hypotenuse from point
P1 to P2 which is absolutly
necessary for the
representation of a leaf image.
2.4 Neuronal
Network
Application
Detection of medical phenomena.
Stock market prediction.
Credit assignment.
Monitoring the condition of
machinery
Engine management
Advantages
Power
Ease of use
automatic
robust
Conclusion
From all this discussion we finally can conclude
that If the proper implementation and at proper
learning is provided to “Neural Networks” they can
replace the simple processors of today.
Neural networks offer the following desirable
qualities:
Automatic search of all possible interrelationships
among key factors
Automatic modeling of complex problems without
prior knowledge of the level of complexity
Ability to extract key findings much faster than
many other tools
Thank You
Back
Back
back
Layers
Back
Learning
 Unsupervised learning .
 Reinforcement learning.
 Back-propagation.
 Online /Offline learning.
 Learning law.
Communication types of
connections
 Intra layer connection

 Recurrent
 On –center /Off surround
Communication types
of connections
 Inter layer connection
 Fully connected
 Partially connected
 Feed forward
 Bi-directional
 Hierarchical
 Resonance

Prepared By: Meera S. Gohil Hansa J. Gajera: V.V.P. Engg. College

Uploaded by

Copyright:

Available Formats

Prepared By: Meera S. Gohil Hansa J. Gajera: V.V.P. Engg. College

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Prepared By: Meera S. Gohil Hansa J. Gajera: V.V.P. Engg. College

Uploaded by

Copyright:

Available Formats

Neural network

V.V.P. ENGG. COLLEGE

The training algorithm for a BPN consists of the following steps:

 Selection and Preparation of Training Data

LeavesRecognition is a neuronal network based java

The intention of this application is to give the user the

2.1 Image edge detection

The Prewitt edge detection filter uses these two 3x3

All pixels are filtered. In order to filter pixels

The used thinning algorithm here

Red Square: This square

Blue line: The compound of

With this two angles we can

 Intra layer connection

You might also like