2. INTRODUCTION
Traditional face recognition algorithms works on a single input image for
feature extraction
Low accuracy
Our implementation works on multiple images
for feature extraction
3. For the system that we have developed we initially extract the features of
the face using Principal Component Analysis and then feed the extracted
Eigen values into the input of a feed forward back-propagation neural
network.
We have taken a data set of 8 persons with 10 images of each person.
The neural network consists of 1 input layer(10 neurons), 3 hidden layers(10
neurons each) and 1 output layer(1 neuron).
4. EIGEN FACE
In Eigen Faces approach, the face images are decomposed into a small set
of characteristic feature images called “Eigen-faces” (which contain the
common features in a face) which are extracted from the original training set
of images by means of principal component analysis. An initial set of images is
acquires(training set) and the Eigen Faces from the training set are
calculated and only M images that correspond to the highest Eigen Values
define the face space.
5. NEURAL NETWORKS
A neural network usually involves a large number of processors operating
in parallel, each with its own small sphere of knowledge and access to
data in its local memory. A neural network is initially "trained" or fed large
amounts of data and rules about data relationships. A program can then
tell the network how to behave in response to an external or can initiate
activity on its own. Neural networks are typically organized in layers. Layers
are made up of a number of interconnected 'nodes' which contain an
'activation function'. Patterns are presented to the network via the 'input
layer', which communicates to one or more 'hidden layers' where the
actual processing is done via a system of weighted 'connections'. The
hidden layers then link to an 'output layer' where the answer is output.
6. CONTROL FLOW CHART
OPEN WEB CAMERA ADD PERSON SAVE EIGEN VALUES IN A
FILE
INPUT EIGEN VALUES TO
NEURAL NETWORK
PASS INPUT VALUES
THROUGH THE NETWORK
ERROR CALCULATION
USING DESIRED OUTPUT
BACK PROPAGATING
CHANGING WEIGHTS
FOR EPOCHS
LESS THAN
ASSIGNED
VALUE
MATCH FACE WITH MOST
NEAR VALUE OR LEAST
ERROR