Pre-Defense CSE Thesis Presentation in BAUST

Handwriting Character Recognition Using Deep Learning
and Computer Vision
Bangladesh Army University of Science and Technology
Tentative Title
Course Code: CSE 4100
Department of Computer Science and Engineering
Course Title: Thesis and Project

Presenter
1. Md.Delwar Hosen Chowdhury(150201002)
2. Horidash Chandro Roy(150201011)
3. Naiyan Noor (150201018)
4. Isfat Zahan Nila (150201020)
Supervisor
Dr. Mohammed Sawket Ali
Asst. Professor
Department of Computer Science & Engineering
Handwriting Character Recognition Using Deep Learning
and Computer Vision
2
2

Presentation Outlines
Introduction
Objective
Outline Methodology Design
 Conversion to Gray-Scale
 Pre-Processing
 Thresholding
Resource Required to Accomplish the Task
 Image Segmentation
Conclusion
 Acquisition from trained character
 Recognized character
3
3

Introduction
Handwriting recognition is the ability of a computer or device to take
input handwriting from sources such as printed physical documents,
pictures and other devices to use handwriting as an input to a
touchscreen and then interpret this as text.
4
4

Objective
 Study; Implementation of the different method used in Computer
Vision and Deep learning to recognize handwritten character.
 To Implement the developed method using python language
 To provide an easy user interface to input the object image
5
5

Outline
Methodology
Design
6
6

 This is the first step performed in image
processing.
 In this step the noise from the image is
removed by using median filtering.
 Median filtering is one of the most widely
used noise reduction technique.
 This is because in median filtering the
edges in image are preserved while the
noise is still removed.
Pre-Processing
Fig: 2 :Pre-Processing
7
7

Conversion to Gray-Scale
 After the pre-processing step, the image
is converted into grayscale.
 Conversion into grayscale is necessary
because different writers use pens of
different color's with varying intensities.
 Also working on grayscale images
reduces the overall complexity of the
system.
Fig:3:Conversion to Gray-Scale
8
8

Thresholding
 When an image is converted into
grayscale, the handwritten text is darker as
compared to its background.
 With the help of thresholding we can
separate the darker regions of the image
from the lighter regions.
 Thus because of thresholding we can
separate the handwritten text from its
background.
Fig:4:Thresholding
9
9

Image Segmentation
 A user can write text in the form
of lines.
 The thresholder image is ﬁrst
segmented into individual lines.
 Then each individual line is
segmented into individual
words.
Fig:5:Image Segmentation
10
10

 Standard Data Set
 Python Language
 Spyder and Jupiter Notebook
Resource Required to Accomplish the Task
11
11

Conclusion
Currently our system will recognize character of English alphabet We try to
add support for more languages in our works one by one. But presently we
try to develop the system which can only recognize handwriting of English
letters and digits After then we try to add support for recognition of cursive
text.
Some of the applications are processing of cheques in Banks, helping hand
in Desktop publishing, Recognition of text from business cards, helping the
blind in recognizing handwritten text on letters etc.
12
12

References
[1] Wei Lu, Zhijian Li,Bingxue Shi . ” Handwritten Digits Recognition with Neural Networks and Fuzzy Logic” in
IEEE International Conference on Neural Networks, 1995. Proceedings.
[2] B. V. S. Murthy.” Handwriting Recognition Using Supervised Neural Networks” in International Joint
Conference on Neural Networks, 1999. IJCNN ’99.
[3] M. Gilloux, J.-M. Bertille, and M. Leroux, “Recognition of Handwritten Words in a Limited Dynamic
Vocabulary,” Third Int’l Workshop Frontiers in Handwriting Recognition, pp. 417–422, CEDAR, State Univ. of
New York at Buffalo, May 1993.
[4] S. Edelman, S. Ullman, and T. Flash, “Reading Cursive Script by Alignment of Letter Prototypes,” Int’l J.
Computer Vision, vol. 5, no. 3, pp. 303–331, 1990.
13
13

Any Question?
Handwriting Recognition Using
Deep Learning and Computer
Vision
14
14

Pre-Defense CSE Thesis Presentation in BAUST

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Pre-Defense CSE Thesis Presentation in BAUST

Similar to Pre-Defense CSE Thesis Presentation in BAUST (20)

More from Naiyan Noor

More from Naiyan Noor (20)

Recently uploaded

Recently uploaded (20)

Pre-Defense CSE Thesis Presentation in BAUST

Editor's Notes