You are on page 1of 24

AMRUTVAHINI COLLEGE OF ENGINEERING, SANGAMNER

Department of Computer Engineering

2021-2022
A
Project presentation
on
“Devanagari Character Recognition Using Deep Learning”

By
Group Id-4111
Mr. Hase Atharva Shivaji(71902512F)
Mr. Jain Viral Vivek(71902532L)
Mr. Gujarathi Parag Rajesh(71902496L)
Ms. Bhor Aditi Ranjeet(71902355G)
Ms. S. B. Bhonde Dr. M. A. Wakchaure
Project Guide Project Coordinator
Dept. of Computer Engineering Dept. of Computer Engineering
Prof. R. L. Paikrao
H.O.D
Dept. of Computer Engineering
Contents
 Introduction
 Problem Statement
 Objectives
 Literature Review
 Scope
 Software Requirements
 System design and Dataset
 Data Flow Diagram
 Class Diagram and Activity Diagram
 System Architecture and methodology of problem solving
 Snapshot of Code
 Applications
 Result
Introduction
• Handwriting recognition is the ability of a machine to receive and interpret
handwritten input from multiple sources like paper documents, photographs ,
touch screen devices etc.

• Word recognition is very important area of Document Analysis and Recognition.

• Convolution neural networks uses data driven learning and extraction of


hierarchical features from training Convolution layers in CNN.

• This system will provide the end users to convert the handwritten text into digital
format.
Problem Definition
•This project takes handwritten Devanagari characters as input.
Process the character, train the algorithm, to recognize the pattern
and modify the character to a beautified version of the input.
• This project is aimed at developing software which will be helpful
in recognizing characters of Devanagari language.
•In this work we will design and implement the system for
handwritten Devanagari characters recognition which will also play
an important role in education sector too.
Objectives
To enable system to pre-process the given input to suppress
the Background and select the features.
To detect text regions, present in the image.
To identify handwritten characters.
To retrieve text present in the image and display them to the
user.
Literature Survey
Sr.No Title of Paper Author Work On Techniques/Models Used

1 Neural based handwritten Harmandhu M, The new sector-based method and ring-based approach,
. character recognition Murali Mohan K.R the combination of these, termed pre classification of characters.
Kumar H. the Fusion method for the
recognition of handwritten English
capital letters.

2 Handwritten Character Chirag 1 Patel, Ripal The behaviours of different Models segmentation of characters,
Recognition using Neural Patel, Palak Patel. of Neural Network used in OCR. normalizing of characters and De-
Network OCR is widespread use of Neural skewing.
Network.

3 Character Recognition Kauleshwar Prasad, recognition of English alphabet in a noise filtering, smoothing and
Using MATLAB’s Neural Devvrat C. Nigam, given scanned text document with normalization, character extraction
Network Toolbox Ashmika Lakhotiya the help of Neural Networks. Using and edge detection algorithm
and DheerenUmre Mat lab Neural Network toolbox.
Scope of Project
•In pre-processing stage, the quality of the input image is checked
& the data of interest is located.
•The feature extraction stage which captures the distinctive
characteristics of the digitized characters for Recognition.
•The classification stage which processes the feature vectors to
identify the characters and words.
•In order to do so the algorithm has to depend on the distributional
properties of the image, this requires a rigorous analysis and
study of particular image and still rigorous programming.
Software Requirements
Specification
ASSUMPTION AND DEPENDENCIES :
•Assumption:
User of the system should be trained.
Pre-Trained Dataset should be provided for proper output.

•Dependencies:
Trained dataset should be proper.
Image processing DLL and Dataset are the dependencies of the project.
Software Quality Attributes
Adaptability
Proposed System can be adapted easily to various Python versions.

Availability
Can easily execute on currently available minimum configuration of
hardware and Software.

Correctness
It will work correctly according to the valid input requirements.

Usability
Can integrate system according to available resources like Windows OS
Versions. Other quality attributes are like portability, maintainability,
reliability, testability.
Hardware Requirements:
Processor : Intel core i3 7th generation with speed 2GHz
Ram : 4GB or more
Hard Disk: 1 TB

Operating System:
Windows7/8/9/10

Technology:
Python-3[PL], DL.
Diagrams
1.DFD Diagrams
• DFD 0
• DFD 1
• DFD 2
2.Class Diagram
3.Use case Diagram
4.Activity Diagram
Data Flow Diagrams
DFD 0:

DFD 1:
DFD 2:
Class Diagram
Activity Diagram
System Architecture
Methodology of Problem
Solving
Snapshot of Code
Applications
• 100% text searchable documents
• It is cheaper than paying someone amount to manually enter
great deal of text data. Moreover it takes less time to convert
within the electronic form.
• It is used in post office for reading postal address.
• To convert the fax and news papers into text format.
• A paper based form are often became an electronic form which
is straightforward to store or send by mail.
Result
References
1.Bappaditya Chakraborty, Bikash Shaw, Jayanta Aich, Ujjwal Bhattacharya and
Swapan Kumar Parui “Does Deeper Network Lead to Better Accuracy: A Case
Study on Handwritten Devanagari Characters” 2018 13th IAPR International
Workshop on Document Analysis Systems. 2018 IEEE DOI 10.1109/DAS.2018.7
2.Rahul Kala, Harsh Vazirani, Anupam Shukla and Ritu Tiwari, ”An Overview of
Character Recognition Focused on OnLine Handwriting”, IEEE Transactions On
Systems, Man, And Cybernetics Part Applications And Reviews, Vol. 31, NO. 2,
MAY 2001. Access Date:09/07/2015.
3.Kumar, Satish. "A study for handwritten Devanagari word recognition."
Communication and Signal Processing (ICCSP), 2016 International Conference
on. IEEE, 2016
4. N. Sankaran and C. V Jawahar, “Recognition of Printed Devanagari Text Using
BLSTM Neural Network,” 21st International Conference on Pattern Recognition
(ICPR 2012), pp. 322–325, 2012
5. P. K. Sonawane, “Handwritten Devanagari Character Classification using Deep
Learning .,” 2018 Int. Conf. Inf. , Commun. Eng. Technol., pp. 1–4, 2018.
Thank You..!!

You might also like