You are on page 1of 14

HANDWRITTEN DIGIT RECONGITION

CONTENTS
• ABSTRACT
• EXISTING SYSTEM
• DISADVANTAGE OF EXISTING SYSTEM
• PROPOSED SYSTEM
• ADVANTAGES OF PROPOSED SYSTEM
• CONTENT DIAGRAMS
• UML DIAGRAMS
• SYSTEM REQUIREMENTS
• CONCLUSION
• REFERENCE
ABSTRACT
• A handwritten digit recognition system was used in a
demonstration project to visualize artificial neural
networks, in particular Kohonen’s self-organizing feature
map. The purpose of this project was to introduce neural
networks through a relatively easy-to-understand
application to the general public. This paper describes
several techniques used for preprocessing the
handwritten digits, as well as a number of ways in which
neural networks were used for the recognition task.
Existing system
• The current systems use resource intensive image
processing methods and algorithms with varying
accuracies and rejections.
• The feature to be extracted should have minimum
variance within a class and maximum variance between
classes.
• The Classifier to be used should able to classify digit
with high accuracy and should take less training time
during classification. This thesis focuses on Offline
Recognition System.
DISADVANTAGES OF EXISTING
SYSTEM
• SVM classifier takes long training time
• As per existing study, KNN and SVM classifiers could
pose scalability difficulties, and could also be a
comparatively unreliable classifier.
Proposed system
• The proposed system uses efficient pattern classification
and machine learning approach to improve overall
performance of predicting the digits.

• Discovering new ways for image classification which are


not generally used when it comes to images. It opens up
whole new realm of possibilities.
ADVANTAGES OF PROPOSED SYSTEM

• Reduces Overfitting redundant data implies less chance


to make decisions based on noise

• Improves Accuracy Less misleading data implies


modeling accuracy improves.

• Reduces Training Time Less data implies that algorithms


train faster.
CONTENT DIAGRAM
UML DIAGRAM
USE-CASE DIAGRAM:
SEQUENCE DIAGRAM:
CLASS DIAGRAM:
Hardware and software requirement

Hardware Requirements:
• Processor: Intel I3 or greater
• RAM: 8GB
• Hard Disk: 50GB

Software Requirements:
• Web Server: Python Flask 1.0.2
• Backend Language: Python 3.6.8
• Database: MNIST dataset
CONCLUSION
• This method to increase handwritten digits recognition rates by
combining feature extractions methods is propose this paper,
using different approaches were extracted and evaluated. The
experiments demonstrated that different feature extraction
algorithms, have better discriminative ability for certain types
digits, therefore they are complementaryed. Six feature sets,
two proposed in.
References
• https://www.python-course.eu/neural_network_mnist.php
• http://yann.lecun.com/exdb/mnist/
• http://www.cs.utoronto.ca/~hinton/absps/pancake.pdf
• https://data-flair.training/blogs/tensorflow-mnist-dataset/

You might also like