You are on page 1of 4

CHARACTER RECOGNITION

Project Report submitted in partial fulfillment of the requirement for the degree of

BACHELOR OF ENGINEERING
IN

INFORMATION SCIENCE

OF

VISVESWARIAH TECHNOLOGICAL UNIVERSITY, BELAGAUM

Submitted by

Anjali K. R.(1DS98IS002) Anurag Jain(1DS98IS003)

Chaitra Nagaraj(1DS98IS008)

Under the guidance of Prashant K Admani

Internal Guide
External Guide

2002

Department of Information Science

DAYANANDA SAGAR COLLEGE OF ENGINEERING, BANGALORE 560078


VISVESWARIAH TECHNOLOGICAL UNIVERSITY, BELAGAUM

DAYANANDA SAGAR COLLEGE OF ENGINEERING, BANGALORE 560078

Department of Information Science

Certificate

This is to certify that the project work entitled Character Recognition

Submitted by

Anjali K. R. (1DS98IS002)

Anurag Jain (1DS98IS003)

Chaitra Nagaraj (1DS98IS008)

is a bonafide record of work carried out by them at __________________ in partial fulfillment of the

requirements for the award of degree of Bachelor of Engineering in Information Science of

VISWESWARIAH TECHNOLOGICAL UNIVERSITY for the academic year 2001-2002.

Guide Head of the Department


SYNOPSIS

The ability to identify machine printed characters in an automated or a semi-automated


manner has obvious applications in numerous fields. Since creating an algorithm with a
one hundred percent correct recognition rate is quite probably impossible in our world of
noise and different font styles, it is important to design character recognition algorithms
with these failures in mind so that when mistakes are inevitably made, they will at least
be understandable and predictable to the person working with the program.

This product explores one such algorithm, which can be tested, on several characters. The
percentage of recognition is shown for each recognized character.

The project is an OCR application that combines both neural networks and other means
to recognize handwritten characters. The characters can be mouse-drawn or it can be
scanned and taken as input to the program.

In combining both neural networks and other means, we are able to show that both
methods are efficient. Both the methods used are capable of learning.

The neural networks application uses a very simple method for its running. Each neuron
is checked to see if it is active; if the neuron is active, then it’s weight is increased;
otherwise the neuron’s weight is decreased.

The other approach also uses a simple approach. The drawn character is compared to a
database that is stored in the form of a binary file and if there is a match, then the
appropriate character is displayed.
The report contains:

An introduction to OCR
OCR history
Handwriting Recognition
OCR using pre-processing algorithms
An introduction to Neural Networks
A “How To” on using the application

REQUIREMENT ANALYSIS:

HARDWARE REQUIREMENTS

PENTIUM II
64 MB RAM
8 GB HDD

SOFTWARE REQUIREMENTS

VISUAL C++
VISUAL BASIC FOR FRONT-END

PLATFORM

WINDOWS 98/2000

IMPLEMENTATION METHOD:

The project will be implemented using Visual C++ and Visual Basic.

FUTURE ENHANCEMENTS
The project can be further developed to recognize words, sentences etc.

You might also like