IRIS RECOGNITION USING
MACHINE LEARNING TECHNIQUE
Under the Guidance of:
Dr.M.Ravikanth
PRESENTED BY:
G. Manish Reddy (2011CS010099)
M. Rahul (2011CS010164)
M. Shivadeep (2011CS010166)
G. Purna Teja (2011CS010127)
CONTENTS
Abstract
Existing System and Disadvantages
Proposed System and Advantages
System Requirements
Architecture
UML Diagrams
Result
Conclusion
ABSTRACT
Iris texture can be considered as a physical password since
it has unique features for every person.
Iris biometric widely used as a system for maintaining data
security,such as cellular,phone,etc.
The project is implemented in MATLAB using mmu-iris-
dataset.
EXISTING SYSTEM
The Existing method presents an Iris image recognition
using the K-NN machine learning algorithm.
Along with machine learning,image processing also
included.
Images are segmented and features are extracted and then
the features are trained using the K-NN algorithm.
DISADVANTAGES
Always needs to determine the value of K which may be complex
some time by using KNN.
The computation cost is high because of calculating the distance
between the data points for all the training samples.
Miss Classification
Less Accuracy
PROPOSED SYSTEM
The proposed method presents an Iris image recognition using the
CNN(Convolutional neural network) algorithm.
Using CASIA IRIS dataset we are training CNN model and then we
use this CNN model to predict or recognize persons.
The extracted features are saved as a dataset and trained by the
considered machine learning algorithm.
ADVANTAGES
High accuracy
No miss classifications
It is capable of handling large datasets with high dimensionality.
It enhances the accurancy of the model and prevents the over
fitting issue.
SYSTEM REQUIREMENTS
Software Requirements
Operating System:Windows 10
Python IDLE:python 3.7.0
Hardware Requirements
Processors:Any Intel or AMD X86-64 processor
Disk:500 GB minimum
RAM:4GB and Higher
ARCHITECTURE
Activity Diagram
Usecase Diagram
Sequence Diagram
RESULT
CONCLUSION
Iris Recognition System will recognize the particular iris we are
giving as input using CNN algorithm. The detected iris is marked for
features extraction.
The proposed method shows more accuracy than the previous
existing methods and has no miss classifications.
