Software Requirements Specification on Bengali Braille to Text Translator

Bengali Braille to Text Translator
Software Requirements Specification and Analysis
25-Sep-2016

Software Requirements Specification and Analysis
Submitted by
Minhas Kamal
Roll: BSSE-0509
Student of BSSE 8th Semester
Institute of Information Technology
University of Dhaka
Supervised by
Dr. Mohammad Shoyaib
Associate Professor
University of Dhaka
Submission Date
25th September, 2016

LETTER OF TRANSMITTAL
25th September, 2016.
BSSE 4th Year Exam Committee
University of Dhaka
Sir,
I have prepared the report on Software Requirements Specification of
Bengali Braille to Text Translator.
The primary purpose of this report is to summaries my findings that I
have gathered during the requirements specification process. This
report also includes details of each step I have followed while collecting
the requirements.
Sincerely Yours,
Minhas Kamal
Roll: BSSE-0509
Student of BSSE 8th Semester
University of Dhaka
Enclosure: Software Requirements Specification Report

i
Executive Summary
Braille is a specialized writing system for visually impaired people,
where raised dots on embossed paper are used as tactile alphabet. The
tool Bengali Braille to Text Translator will take in scanned image of
Bengali Braille writing, apply pattern recognition, and translate it to
text. The user does not need much theoretical or technical skill to run
this software.
Acknowledgements
By the grace of Almighty Allah I have completed my report on Software
Requirements Specification of Bengali Braille to Text Translator.
I am grateful to my supervisor Dr. Mohammad Shoyaib for his direction
throughout the working time. It was almost impossible for me to
complete this SRS document without him.
I am also thankful to the teachers and students of The Institute of
Education and Research, University of Dhaka. They greatly helped me in
collecting information among all business.

ii
l to the Program Coordinators of PGDIT. They greatly helped us
collecting information among all business.
Table of Contents
Chapter 1: Introduction ............................................................................ 1
1.1 Purpose 1
1.2 Intended Audience 1
Chapter 2: Inception .................................................................................. 3
2.1 Introduction 3
2.1.1 Identifying Stakeholders 3
2.1.2 Asking the First Questions 4
2.1.3 Recognizing Multiple Viewpoints 4
2.1.4 Working towards Collaboration 5
2.2 Conclusion 6
Chapter 3: Elicitation ................................................................................ 7
3.1 Introduction 7
3.2 Eliciting Requirements 7
3.3 Collaborative Requirements Gathering 7
3.4 Quality Function Deployment 8
3.4.1 Normal Requirements 8
3.4.2 Expected Requirements 8
3.4.3 Exciting requirements 9
3.5 Usage Scenarios 9
3.6 Elicitation Work Product 10
Chapter 4: Scenario-Based Model ........................................................ 11
4.1 Introduction 11
4.2 Use Case Scenario 11
4.3 Use Case Descriptions 13
4.3.1 Bengali Braille to Text Translator 14
4.3.1.2 Pre-Processing 18
4.3.1.3 Translation 31
4.3.1.4 Post-Processing 41

iii
Table of Contents
Chapter 5: Data Model .............................................................................. 48
5.1 Introduction 48
5.2 Data Object Selection 48
5.3 Data Objects & Attributes 51
5.4 Conclusion 52
Chapter 6: Class-Based Model ............................................................... 53
6.1 Introduction 53
6.2. General Classification 53
6.3 Selection Characteristics 55
6.4 Attribute Selection 56
6.5 Defining Methods 57
6.5.1 Verb List 57
6.5.2 Selected Methods 58
6.6 Class Diagram 60
6.7 Class Card 61
Chapter 7: Behavioral Model .................................................................. 62
8.1 Introduction 62
8.2 Identifying Events 62
8.3 State Transition Diagram 64
8.4 Sequence Diagram 65
Chapter 8: Conclusion .............................................................................. 76
Appendix ........................................................................................................ 77

iv
List of Figures
Figure No. Figure Name Page No.
Figure 4.3 Use Case Diagram of BBTT (Level-0) 13
Figure 4.3.1 Use Case Diagram of BBTT (Level-1) 14
Figure 4.3.1.1a Activity Diagram- Image Acquisition 16
Figure 4.3.1.1b Swim-Lane Diagram- Image Acquisition 17
Figure 4.3.1.2 Use Case Diagram of Pre-Processing (Level-1.2) 18
Figure 4.3.1.2.1a Activity Diagram- Enhancement 20
Figure 4.3.1.2.1b Swim-Lane Diagram- Enhancement 21
Figure 4.3.1.2.2a Activity Diagram- Noise Reduction 23
Figure 4.3.1.2.2b Swim-Lane Diagram- Noise Reduction 24
Figure 4.3.1.2.3a Activity Diagram- Connectivity Improvement 26
Figure 4.3.1.2.3b Swim-Lane Diagram- Connectivity Improvement 27
Figure 4.3.1.2.4a Activity Diagram- Quantization 29
Figure 4.3.1.2.4b Swim-Lane Diagram- Quantization 30
Figure 4.3.1.3 Use Case Diagram of Translation (Level-1.3) 31
Figure 4.3.1.3.1a Activity Diagram- Segmentation 33
Figure 4.3.1.3.1b Swim-Lane Diagram- Segmentation 34
Figure 4.3.1.3.2a Activity Diagram- Pattern Recognition 36
Figure 4.3.1.3.2b Swim-Lane Diagram- Pattern Recognition 37
Figure 4.3.1.3.3a Activity Diagram- Raw Text Generation 39
Figure 4.3.1.3.3b Swim-Lane Diagram- Raw Text Generation 40
Figure 4.3.1.4 Use Case Diagram of Post-Processing (Level-1.4) 41
Figure 4.3.1.4.1a Activity Diagram- Spell Check 43
Figure 4.3.1.4.1b Swim-Lane Diagram- Spell Check 44
Figure 4.3.1.4.2a Activity Diagram- Clean and Generate Output 46
Figure 4.3.1.4.2b Swim-Lane Diagram- Clean and Generate Output 47
Figure 6.6 Class Diagram (BBTT) 60
Figure 7.3 State Transition Diagram- User 64
Figure 7.4.1 Sequence Diagram- Take Image 65
Figure 7.4.2 Sequence Diagram- Run Enhancement 66
Figure 7.4.3 Sequence Diagram- Reduce Noise 67
Figure 7.4.4 Sequence Diagram- Improve Connectivity 68
Figure 7.4.5 Sequence Diagram- Apply Quantization 69
Figure 7.4.6 Sequence Diagram- Extract Braille Pattern 70

v
List of Figures
Figure No. Figure Name Page No.
Figure 7.4.7 Sequence Diagram- Convert Braille to Code 71
Figure 7.4.8 Sequence Diagram- Map Code to Character 72
Figure 7.4.9 Sequence Diagram- Check Spell 73
Figure 7.4.10 Sequence Diagram- Run Cleaning Procedure 74
Figure 7.4.11 Sequence Diagram- Customize Template 75

Software Requirement Specification and Analysis

1
Chapter 1
Introduction
1.1 Purpose
This is the Software Requirements Specification (SRS) for the tool- Bengali Braille
Character Recognizer. The document contains detailed functional, non-functional,
and support requirements for the project. It also establishes a requirements
baseline for the development of the system.
The SRS serves as the official means of communicating user requirements to the
developer [1]. It provides a common reference point for both the developer team
and stakeholder community. This document’s content may evolve over time as
users and developers work together in developing the system. But its
fundamental output will remain unchanged.
1.2 Intended Audience
This SRS is intended for all the stakeholders- customers, project managers,
designers, developers, and quality assurance engineers [1] [2]. They will use it for
the following reasons-
o Customers will use this SRS to verify acceptability of the development
team’s work.
o The project managers will design their plan on the basis of this document.
They will set milestones and delivery dates. They will also ensure that the
development team is on track.
o The designers will prepare the systems design based on this document.
They will continually refer back to this SRS to ensure that the system they
are designing will fulfill the customers’ needs.

2
o Developers will use this document as a basis for developing the system’s
functionality. The developers will link the requirements defined in this SRS
to the software they create to ensure that they have created software that
will fulfill all of the customers’ documented requirements.
o Quality assurance engineers will use the SRS to derive test plans and test
cases. When portions of the software are complete, they will run their tests
on that software to ensure that the software fulfills the requirements
documented in this SRS. They will again run their tests on the entire system
when it is complete and ensure that all requirements documented in this
SRS have been completed.

3
Chapter 2
Inception
2.1 Introduction
Requirements Engineering starts with Inception phase. Its goal is to identify
parallel needs and conflicting requirements among the stakeholders of a project.
The foundation was established by following the subsequent factors-
2.1.1 Identifying Stakeholders
Any person, group, or organization which will affect or be affected by the system
directly or indirectly is a stakeholder. It includes both project developers and end-
users. Here only client side stakeholders will be focused.
Although, we intend to develop Bengali Braille to Text Translator for public use,
we are currently building it for using only in the Institute of Education and
Research (IER), University of Dhaka premises. For this reason, we have selected
the stakeholders from the scope of IER only. We have identified following
stakeholders for our project:
1. Teachers of IER: Teachers working with visually impaired student at IER
are our primary client side stakeholders. They will directly interact with
the system.
2. Visually Impaired Students of IER: Although the visually impaired
students will not interact with the system directly, but they are the
biggest group affected by the system. The software will process their
writing and convert it to text.
3. General Students of IER: Visually normal students will also use the
software. But they will not be using it for the same reason as the
teachers.

4
2.1.2 Asking the First Questions
First set of context-free questions focuses on the client side stakeholders and tries
to reveal overall project goals and benefits. Now we ask question to help us
gaining a better understanding of problem, and to allow customers to voice their
perceptions about the solution.
2.1.3 Recognizing Multiple Viewpoints
We have collected these view points by discussing with the teachers, visually
impaired students, and general students of IER.
1. Teachers of IER:
a. Full control of the system
b. Low cost
c. Minimum error rate
d. Handing batch of images
e. Intuitive user interface
f. Ability to see intermediate results
2. Visually Impaired Students of IER:
a. Error free solution
b. Process double sided writing
c. Special character recognition ability
d. Fast processing
3. General Students of IER:
a. Easy to use
b. User friendly interface

5
c. Easy to install
d. Work with any scanner
e. No need of theoretical or technical knowledge
2.1.4 Working towards Collaboration
Every stakeholder has his own set of requirements from his point of view. We
followed following steps to merge these requirements:
i. Identify the common and conflicting requirements
ii. Categorize the requirements
iii. Take priority points for each requirement from stakeholders and on the
basis of this voting prioritize the requirements
iv. Make final decision about the requirements
Common Requirements:
 User friendly interface
 Easy to use
Conflicting Requirements:
 Low cost and process double sided writing
 Fast and minimum error rate
 Full control of the system and no need of theoretical or technical
knowledge

6
Final Requirements: We have finalized following requirements for the system
through categorization and prioritization process:
 Easy to use
 User friendly interface
 Users with theoretical and technical knowledge will be able to fully
customize and control the system
 Users with no technical knowledge will be able to use templates (created
by specialists) for running the system
 Minimize error rate
 Can work with any scanner
2.2 Conclusion
Inception phase strongly supported us in establishing basic understanding about
Bengali Braille to Text Translator. It also helped us to identify people who will be
benefited by the software. Most importantly it established a preliminary
communication with the stakeholders.

7
Chapter 3
Elicitation
3.1 Introduction
Elicitation helps the customer to define the requirement more specifically. This
phase faces many problems like- problems of scope, problems of volatility, and
problems of understanding. To overcome these problems, we have worked in an
organized and systematic manner.
3.2 Eliciting Requirements
Unlike Inception, where Question and Answer approach is used, Elicitation makes
use of a requirements elicitation format that combines the elements of problem
solving, elaboration, negotiation, and specification. It requires the cooperation of
a group of end-users and developers to elicit requirements.
3.3 Collaborative Requirements Gathering
There are many different approaches to collaborative requirements gathering.
Each approach makes use of a slightly different scenario. We followed the
subsequent steps to do it:
i. Meetings were conducted with teachers and students of IER, DU. They
were questioned about their requirements and expectations from the tool.
ii. They were asked about the problems they are facing with exam papers
written in Braille. We also inquired regarding the efficiency of the current
process.
iii. At last we selected our final requirement list from these meetings.

8
3.4 Quality Function Deployment
The technique which translates the needs of the customer into technical
requirements for software is called Quality Function Deployment (QFD).
QFD concentrates on maximizing customer satisfaction from the Software
Engineering process. With respect to our project the following requirements are
identified by QFD-
3.4.1 Normal Requirements
Objectives and goals that are stated during the meeting with the customers are
Normal Requirements. Our project’s Normal Requirements are:
1. User can customize the work sequence, and store the configuration for
future use.
2. Will be able to run the tool and get output with only single ‘run’ button.
3. User can view result from each level of Braille to text translation
process.
4. User will be able to run different types of image enhancement and other
pre-processing algorithms to improve input image quality.
5. There will be a spell checking system, which will minimize error rate.
6. There will be some default configuration templates available with the
software.
7. User will also be able to select different pattern recognition algorithms
for recognizing Braille patterns on the basis of his need.
3.4.2 Expected Requirements
Expected Requirements are implicit to the system. These requirements may be so
fundamental that the customers do not explicitly state them. Expected
Requirements of our project are:

9
1. The system will be able to support all popular image formats like- JPG,
JPEG, PNG, and so on.
2. User will be able to process both single image and batch of images of
Braille writing.
3. The user interface of the system shall be easy to use. It will make use of
drop-down boxes, radio buttons, and other selectable fields wherever
possible instead of fields that require the user to type in data.
3.4.3 Exciting Requirements
Features that go beyond the customer's expectations are called Exciting
Requirements. They prove to be very satisfying when present-
1. The user interface should provide appropriate error messages for invalid
input as well as tool-tips and help.
2. The system will also be able to work with multiple languages of Braille
writing.
3.5 Usage Scenario
Bengali Braille Character Recognizer is a tool that will take scanned image of
Braille writing, run different types of image-preprocessing techniques, translate
Braille to text through pattern recognition, and apply text correction procedures
for final output.
Braille is a specialized writing system for visually impaired people. Here raised
dots on embossed paper are used as tactile alphabet [3] [4]. These papers will be
scanned with a scanner, and resulting image files will be stored in the computer.
At first, the user will select scanned images of Braille writing. Now, different types
of image enhancement processes will be run. These image enhancement
processes will improve Braille patterns and make it distinguishable from the

10
background. After that, various noise reduction algorithms will be applied for
reducing artifacts and improve pattern quality. For further development of Braille
dots, connectivity improvement and quantization will be applied.
In the second section, Braille is translated to text. Firstly, Braille patterns are
extracted from the image. Then these patterns are converted to code- Braille
Pattern Code (BPC- each dot of Braille character is encoded in ‘1’ and ‘0’). Now
these resulting codes are mapped with Code to Character Map (CCM-like a hash
table, where each unique code relates to only one character) and plain text file is
generated. This text file contains probable raw characters for the Braille image.
In the post processing stage, spell checking will be performed for minimizing error
rate. Here, words for spell checking process will be stored in a Domain Word List
(DWL) file. Lastly, a cleaning procedure will produce the final text file.
All these steps, settings, algorithms and their parameters can be automated using
a configuration file- Braille Processing Template (BPT). Some templates (BPT) will
be available to the user in default with the system. Users will also be able to
define customized template.
3.6 Elicitation Work Product
Our elicitation work product includes:
 Statement of our requirements for Bengali Braille to Text Translator
 A bounded statement of scope for our solution
 Set of usage scenarios
 A list of clients, users, and other stakeholders who participated in
requirement specification process

11
Chapter 4
Scenario-Based Model
4.1 Introduction
The user’s point of view is used to describe Scenario-Based Model. In SRS this is
the first modeling phase. So, it serves as an input for the creation of other models.
4.2 Use Case Scenario
With the advancement of requirements gathering, functionalities and
responsibilities of the software starts to materialize. The following table enlists
primary components of the system:
Table-4.2 Use Case Scenario
Level-0 Level-1 Level-2 Actors
Bengali
Braille to
Text
Translator
Pre-Processing Image Acquisition User, Scanner
Enhancement User, Braille Processing
Template
Noise Reduction User, Braille Processing
Template
Connectivity
Improvement
User, Braille Processing
Template
Quantization User, Braille Processing
Template
Translation Segmentation User, Braille Processing
Template
Pattern Recognition User, Braille Processing
Template, Braille

12
Pattern Code
Raw Text Generation User, Braille Processing
Template, Code to
Character Map
Post-
Processing
Spell Check User, Braille Processing
Template, Domain
Word List
Clean and Generate
Output
User, Braille Processing
Template

13
4.3 Use Case Description
In this section Use Case Scenario will be elaborated to Use Case Diagram,
Description, Activity Diagram, and Swim-Lane Diagram. Figure-4.3 is the Use Case
Diagram of level-0 for Bengali Braille to Text Translator (BBTT):

14
4.3.1 Bengali Braille to Text Translator
Here Figure-4.3.1 is the detailed form of level-0:

15
4.3.1.1 Use Case: Image Acquisition
Primary Actor: User.
Secondary Actor: Scanner, Braille Processing Template.
Goal in Context: Use scanner to scan Braille writing and initialize the
system.
Scenario:
1. Place Braille writing sheets in scanner.
2. Scan and store images.
3. Input images into the system.
4. Select Braille Processing Template (BPT) if
necessary.
Exceptions:
1. System failure.
2. Error in scanner connection.
Priority: Moderate, may be implemented.
When Available: First increment.
Frequency of Use: Several times per week.

16
Figure 4.3.1.1a is the Activity Diagram of Image Acquisition-

17
Figure 4.3.1.1b is the Swim-Lane Diagram of Image Acquisition-

18
4.3.1.2 Pre-Processing
Pre-Processing phase can be divided into five sub-phases. Figure-4.3.1.2 shows
this:

19
4.3.1.2.1 Use Case: Enhancement
Secondary Actor: Braille Processing Template.
Goal in Context: Improve Braille patterns and make it distinguishable
from the background.
Scenario:
1. Select image enhancement algorithm.
2. Set parameters.
3. Run process and store result.
Exceptions:
1. System failure.
2. Image file size is too large.
3. Unsupported image file format.
When Available: Second increment.
Frequency of Use: Several times per day.

20
Figure 4.3.1.2.1a is the Activity Diagram of Enhancement-

21
Figure 4.3.1.2.1b is the Swim-Lane Diagram of Enhancement -

22
4.3.1.2.2 Use Case: Noise Reduction
Goal in Context: Reducing artifacts and improve pattern quality.
Scenario:
1. Select noise reduction algorithm.
2. Set parameters.
Exceptions:
1. System failure.

23
Figure 4.3.1.2.2a is the Activity Diagram of Noise Reduction-

24
Figure 4.3.1.2.2b is the Swim-Lane Diagram of Noise Reduction-

25
4.3.1.2.3 Use Case: Connectivity Improvement
Goal in Context: Improve Braille writing patterns.
Scenario:
1. Select algorithm.
Exceptions:
1. System failure.
2. Image file size is too large.

26
Figure 4.3.1.2.3a is the Activity Diagram of Connectivity Improvement-

27
Figure 4.3.1.2.3b is the Swim-Lane Diagram of Connectivity Improvement-

28
4.3.1.2.4 Use Case: Quantization
Goal in Context: Improve Braille writing patterns.
Scenario:
1. Set parameters.
Exceptions:
1. System failure.

29
Figure 4.3.1.2.4a is the Activity Diagram of Quantization-

30
Figure 4.3.1.2.4b is the Swim-Lane Diagram of Quantization-

31
4.3.1.3 Translation
Further section of Translation system creates three sub-systems:

32
4.3.1.3.1 Use Case: Segmentation
Goal in Context: Extract Braille patterns from the image.
Scenario:
1. Select segmentation procedure.
2. Set parameters.
3. Run process.
4. Store resulting binary image.
Exceptions:
1. System failure.
3. Big image file.
Priority: Essential, must be implemented.
When Available: Third increment.

33
Figure 4.3.1.3.1a is the Activity Diagram of Segmentation-

34
Figure 4.3.1.3.1b is the Swim-Lane Diagram of Segmentation-

35
4.3.1.3.2 Use Case: Pattern Recognition
Secondary Actor: Braille Processing Template, Braille Pattern Code.
Goal in Context: Convert Braille patterns to code.
Scenario:
1. Select pattern recognition procedure.
2. Set parameters.
3. Run process.
4. Store code in text file.
Exceptions:
1. Image with no Braille pattern.
2. System failure.
3. Big image file.
When Available: Fourth increment.

36
Figure 4.3.1.3.2a is the Activity Diagram of Pattern Recognition-

37
Figure 4.3.1.3.2b is the Swim-Lane Diagram of Pattern Recognition-

38
4.3.1.3.3 Use Case: Raw Text Generation
Secondary Actor: Braille Processing Template, Braille Pattern Code,
Code to Character Map.
Goal in Context: Generate plain text file.
Scenario:
1. Select Braille Pattern Code file.
2. Select Code to Character Map file.
3. Run process.
4. Store result in text file.
Exceptions:
1. Invalid pattern code.
2. System failure.
When Available: Fifth increment.

39
Figure 4.3.1.3.3a is the Activity Diagram of Raw Text Generation-

40
Figure 4.3.1.3.3b is the Swim-Lane Diagram of Raw Text Generation-

41
4.3.1.4 Post-Processing
Post-Processing has two major phases:

42
4.3.1.4.1 Use Case: Spell Check
Secondary Actor: Braille Processing Template, Domain Word List.
Goal in Context: Minimizing error rate.
Scenario:
1. Select raw text file.
2. Select Domain Word List file.
3. Run process.
4. Store corrected result in text file.
Exceptions:
1. System failure.
When Available: sixth increment.

43
Figure 4.3.1.4.1a is the Activity Diagram of Spell Check-

44
Figure 4.3.1.4.1b is the Swim-Lane Diagram of Spell Check-

45
4.3.1.4.2 Use Case: Clean and Generate Output
Goal in Context: Generate final text file.
Scenario:
1. Select raw text file.
2. Run cleaning process.
3. Store result in text file.
Exceptions:
1. System failure.
When Available: Last increment.

46
Figure 4.3.1.4.2a is the Activity Diagram of Clean and Generate Output-

47
Figure 4.3.1.4.2b is the Swim-Lane Diagram of Clean and Generate Output-

48
Chapter 5
Data Model
5.1 Introduction
When software requires interfacing with a database or complex data structures
need to be constructed and manipulated, data model is performed as part of
overall requirements modeling.
5.2 Data Object Selection
Data objects are representation of information which has different
attributes.Table-5.2 enlists all nouns from Usage Scenario and marks potential
data objects:
Table-5.2 Data Object Selection
Noun Attributes Description Remark
bengali braille
character
recognizer
represents the whole
system
rejected
image alias of braille writing rejected
braille writing input for the system rejected
image-
preprocessing
technique
alias of algorithm rejected
text alias of text file rejected
pattern recognition alias of algorithm rejected
final output output for the system rejected
braille alias of braille writing rejected
system alias of bengali braille rejected

49
character recognizer
visually impaired
people
out of scope rejected
raised dots out of scope rejected
embossed paper out of scope rejected
tactile alphabet out of scope rejected
paper alias of embossed paper rejected
scanner a device, external entity rejected
image files alias of braille writing rejected
computer represents the whole
system
rejected
scanned images alias of braille writing rejected
image
enhancement
process
section out of scope rejected
braille pattern alias of braille writing rejected
background out of scope rejected
noise reduction
algorithm
artifact out of scope rejected
pattern quality out of scope rejected
braille dots alias of braille writing rejected
connectivity
improvement
quantization alias of algorithm rejected
code an attribute of braille
pattern code
rejected
braille pattern
code
id, code potential data object accepted
braille character alias of braille writing rejected
code to character
map
code, character potential data object accepted
hash table alias of braille code to
character map
rejected
text file alias of final output rejected
braille image alias of braille writing rejected

50
stage alias of algorithm rejected
spell checking alias of algorithm rejected
error rate out of scope rejected
domain word list word, frequency potential data object accepted
cleaning procedure alias of algorithm rejected
step alias of algorithm rejected
setting alias of braille
processing template
rejected
algorithm an attribute of braille
processing template
rejected
parameter an attribute of braille
processing template
rejected
configuration file alias of braille
processing template
rejected
braille processing
template
algorithm,
parameters
potential data object accepted
template alias of braille
processing template
rejected
user no separated
authentication system
required
rejected
system alias of braille writing rejected
customized
template
alias of bengali braille
rejected

51
5.3 Data Objects and Attributes
This is a brief view of all attributes we have found so far:

52
5.4 Conclusion
Data objects found here have no relationship to each other. These data objects
will be stored in a file system and will work independently. So there is no Data
Object Relational Diagram here. For this same reason no E-R Diagram or Schema
Diagram exists for this project.

53
Chapter 6
Class-Based Model
6.1 Introduction
The objects that the system will manipulate, the operations that will be applied to
the objects and relationships between the objects are represented by Class-Based
Modeling. It also defines the collaborations that occur between the classes.
6.2 General Classification
Analysis classes can be marked by one of the following ways:
1. External Entity
2. Thing
3. Occurrence
4. Role
5. Organizational Unit
6. Place
7. Structure
Table-6.2 General Classification
No. Noun General
Classification
Remark
1 bengali braille
2 Problem Space (represents
whole system)
2 image 2, 7 Problem Space
3 braille writing NULL Problem Space
4 image-preprocessing
technique
3, 7 Solution Space (same as- 25)

54
5 text 2, 7 Problem Space
6 pattern recognition 3 Solution Space
7 final output 7 Problem Space
8 braille Null Problem Space
9 system 2 Problem Space (represents
whole system)
10 visually impaired
people
1 Problem Space
11 raised dots 1 Problem Space
12 embossed paper 1 Problem Space
13 tactile alphabet 1 Problem Space
14 paper 1 Problem Space
15 scanner 1, 2 Problem Space
16 image files 1, 2 Problem Space
17 computer 1, 2, 7 Problem Space (represents
whole system)
18 scanned images 1, 2 Problem Space
19 image enhancement
process
3 Solution Space
20 section NULL Problem Space
21 braille pattern 1, 2 Problem Space
22 background NULL Problem Space
23 noise reduction
algorithm
3 Solution Space
24 artifact NULL Problem Space
25 pre-processing 3 Solution Space
26 braille dots 1, 2 Problem Space
27 connectivity
improvement
3 Solution Space
28 quantization 3 Solution Space
29 code NULL Problem Space
30 translation 3 Solution Space
31 braille character NULL Problem Space
32 code to character
map
4, 7 Solution Space
33 hash table 4, 7 Solution Space (same as- 32)
34 text file 1, 2 Problem Space

55
35 braille image 1, 2 Problem Space
36 post-processing 3 Solution Space
37 spell checking 3 Solution Space
38 error rate NULL Problem Space
39 domain word list 4, 7 Solution Space
40 cleaning procedure 3 Solution Space
41 step NULL Problem Space
43 setting 4, 7 Solution Space (same as- 47)
44 algorithm 3 Solution Space (represents all
procedures)
45 parameter 7 Solution Space (represents all
parameters of the system)
46 configuration file 4, 7 Solution Space (same as- 47)
47 braille processing
template
4, 7 Solution Space
48 template 4, 7 Solution Space (same as- 47)
49 user 1, 4 Solution Space
50 system 2 Problem Space (represents
whole system)
51 customized
template
4, 7 Solution Space (same as- 47)
6.3 Selection Characteristics
Coad and Yourdon suggest six selection characteristics that should be used to
consider each potential class for inclusion in the analysis model:
1. Retained Information
2. Needed Services
3. Multiple Attributes
4. Common Attributes
5. Common operations
6. Essential Requirements

56
6.4 Attribute Selection
Here we find attributes for selected classes.
Table-6.4 Attribute Selection
No. Class Attributes
1 User selected_algorithm_List, initialized_parameters
2 Pre-Processing image_input_file_list, output_image_path_list
3 Translation image_list, code_data, text_data,
Table-6.3 Selection Characteristics
No. Potential Class Characteristics Remarks
1 pattern recognition 1, 2, 3, 6 No
2 image enhancement process 1, 2, 3 No
3 noise reduction algorithm 1, 2, 3 No
4 pre-processing 1, 2, 3, 4, 5, 6 Yes
5 connectivity improvement 1, 2, 3 No
6 quantization 1, 2, 3 No
7 translation 1, 2, 3, 4, 5, 6 Yes
8 code to character map 1, 2, 6 No
9 post-processing 1, 2, 3, 4, 5, 6 Yes
10 spell checking 1, 2, 3 No
11 domain word list 1, 2 No
12 cleaning procedure 1, 2, 3 No
13 braille processing template 1, 2 No
14 user 1, 2, 3, 4, 5, 6 Yes

57
code_to_character_map
4 Post-Processing Input_text_data, output_file_path
6.5 Defining Methods
In this section we find all the verbs from usage scenario and include necessary
external verbs in a list. Then we select useful verbs as methods.
6.5.1 Verb List
Here we list all verbs from usage scenario.
Table-6.5.1 Verb List
No. Verb Remarks
1 take scanned image out of scope
2 run image pre-processing yes
3 translate braille to text yes
4 apply text correction yes (same as- run post-
processing)
5 scan paper out of scope
6 store image out of scope
7 select image out of scope
8 run image enhancement yes
9 improve braille pattern yes (same as- run image
enhancement)
10 make distinguishable yes (same as- run image
enhancement)
11 apply noise reduction yes
12 improve connectivity yes
13 run quantization yes

58
14 extract braille pattern yes
15 convert braille pattern to code yes
16 map code to character yes (same as- convert
braille pattern to code)
17 generate plain text from code yes
18 contain character out of scope
19 run post processing yes
20 run spell checking yes
21 store word out of scope
22 run cleaning procedure yes
23 automate whole process no need
24 customize braille processing template yes
6.5.2 Selected Methods
From the verb list above we have selected following methods for classes.
Table-6.5.2 Selected Methods
Class Methods Remarks
User pre_process()
translate()
post_process()
customize_braille_processing
_template()
pre_process() is run
image pre-processing,
translate() is translate
braille to text,
post_process() is run
post-processing
Pre-Processing enhance_image()
reduce_noise()
improve_connectivity()
run_quantization()
enhance_image() is
run image
enhancement,
reduce_noise() is
apply noise reduction
Translation apply_segmentation()
recognize_braille_pattern()
apply_segmentation()
is extract braille

59
create_plain_text() pattern,
recognize_braille_patt
ern() is convert braille
pattern to code,
create_plain_text() is
generate plain text
from code
Post-
Processing
check_spell()
generate_final_output()
generate_final_output
() is run cleaning
procedure

60
6.6 Class Diagram
We have shown here (Figure- 6.6), how the classes interact together to
accomplish certain goal.

61
6.7 Class Card
Class card represents a graphical view of responsibility and collaborator for each
class.

62
Chapter 7
Behavioral Model
7.1 Introduction
When the system is perceived in terms of states and transitions, it is called
Behavior Modeling. It is also known as State Modeling, State Machines, or State
Transition Matrix.
This requires both identifying all of the interesting states of being that software or
its components are likely to be in. And also, at a high level, abstracting what
events are likely to cause software or its components to change between states of
being.
7.2 Identifying Events
Here we have identified events from the Usage Scenario and listed their
corresponding initiators & collaborators.
Table-7.2 Identifying Events
Event Initiator Collaborator
take scanned image of Braille
writing
User Scanner
run different types of image-
preprocessing techniques
User
translate Braille to text
through pattern recognition
User Braille Pattern Code, Code
to Character Map
apply text correction
procedures
User Domain Word List

63
select scanned images of
Braille writing
User
run image enhancement User
apply noise reduction User
improve connectivity User
apply quantization User
extract braille pattern User
convert braille to code User Braille Pattern Code
map code to character User Braille Pattern Code, Code
to Character Map
check spell User Domain Word List
run cleaning procedure User
automate the configuration User Braille Processing Template
customize template User Braille Processing Template

64
7.3 State Transition Diagram
State Transition Diagram represents active states for each class and the events
(triggers) that cause changes between these active states. Here Figure-7.3
provides diagram for the actor- User.

65
7.4 Sequence Diagram
The way events cause transitions from object to object as a function of time is
represented by Sequence Diagram. Figure-7.4.1 to Figure-7.4.11 represents
Sequence Diagram for all major events of this project.

76
Chapter 8
Conclusion
I am pleased to submit the final SRS report on Bengali Braille to Text
Translator. From this, the readers will get a clear and easy view of
tactile writing system. He will also get a good understanding of the
translation process.
This SRS document can be used effectively to maintain the software
development cycle for the project. I have presented a detailed
description of the total system. It will be much easy to conduct the
whole project using this SRS. It will also help me to determine the
pitfalls that may come ahead. Hopefully, this document can also
help other software engineering students as well as practitioners.
I have tried my best to make effective and fully designed SRS. I
believe that the readers will find it in order.

77
Appendix
References
[1] Pressman, Roger S. Software Engineering: A Practitioner's Approach
(7th ed.). Boston, Mass: McGraw-Hill. ISBN 0-07-285318-2.
[2] Ralph, Paul. "The illusion of requirements in software development."
Requirements Engineering 18.3 (2013): 293-296.
[3] W. David, A. Adler. “A Picture of Louis Braille.” New York, McGraw
Hill, 1999.
[4] Durre, K.P., W. Tuttle. “A Universal Computer Braille Code for
Literacy and Scientific Texts.”International Technology Conference, 1991.

Software Requirements Specification on Bengali Braille to Text Translator

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (20)

Similar to Software Requirements Specification on Bengali Braille to Text Translator

Similar to Software Requirements Specification on Bengali Braille to Text Translator (20)

More from Minhas Kamal

More from Minhas Kamal (11)

Recently uploaded

Recently uploaded (20)

Software Requirements Specification on Bengali Braille to Text Translator