SMART SCAN

Capstone Project Report

MID SEMESTER EVALUATION

Submitted by:
(102103278) Chirag Gupta
(102103281) Ishaan Gaba
(102103284) Abhinav Maheshwari
(102103354) Shrey Pachauri
(102153027) Geetika Arora

BE Third Year, CoE

CPG No. 239

Under the Mentorship of

Dr. Smita Agrawal

Assistant Professor

Computer Science and Engineering Department

Thapar Institute of Engineering and Technology, Patiala
July 2024
 ABSTRACT

This project presents an automated marks recognition system designed to enhance the efficiency and accuracy
of capturing and processing marks from scanned documents. The system automates the traditionally manual
process, providing a seamless and reliable solution for educational assessments. By leveraging advanced
methods for object detection and image processing, it ensures precise recognition and handling of marks. The
user interface is designed for ease of use, offering a responsive and intuitive experience. Additionally, the
system supports real-time data management and secure storage, making it scalable and robust. Overall, this
project aims to streamline the assessment workflow, reducing errors and saving time in educational
environments.

2
 DECLARATION

We hereby declare that the design principles and working prototype model of the project entitled Smart Scan
is an authentic record of our own work carried out in the Computer Science and Engineering Department,
TIET, Patiala, under the guidance of Dr Smita Agrawal during 6th semester (2024).

Date: 20th August 2024

Roll No. Name Signature

102103278 CHIRAG GUPTA

102103281 ISHAAN GABA

102103284 ABHINAV MAHESHWARI

102103354 SHREY PACHAURI

102153027 GEETIKA ARORA

Counter Signed By:

Faculty Mentor:

Dr. Smita Agrawal

Assistant Professor

CSED,

TIET, Patiala

3
 ACKNOWLEDGEMENT

We would like to express our thanks to our mentor Dr. Smita Agrawal. She has been of great help in our
venture and an indispensable resource of technical knowledge. She is truly an amazing mentor to have.

We are also thankful to Shalini Batra, Head, Computer Science and Engineering Department, the entire
faculty and staff of the Computer Science and Engineering Department, and also our friends who devoted
their valuable time and helped us in all possible ways towards successful completion of this project. We
thank all those who have contributed either directly or indirectly towards this project.

Lastly, we would also like to thank our families for their unyielding love and encouragement.

They always wanted the best for us, and we admire their determination and sacrifice.

Date: 20th August 2024

Roll No. Name Signature

102103278 CHIRAG GUPTA

102103281 ISHAAN GABA

102103284 ABHINAV MAHESHWARI

102103354 SHREY PACHAURI

102153027 GEETIKA ARORA

4
 TABLE OF CONTENTS

ABSTRACT………………………………………………………………………………….i
DECLARATION……………………………………………………………………………ii
ACKNOWLEDGEMENT………………………………………………………………....iii
LIST OF FIGURES………………………………………………………………………...iv
LIST OF TABLES…………………………………………………………………………..v
LIST OF ABBREVIATIONS……………………………………………………………...vi

CHAPTER……………………………………………………………………………….Page No.
1. Introduction 1
1.1 Project Overview
1.2 Need Analysis
1.3 Research Gaps
1.4 Problem Definition and Scope
1.5 Assumptions and Constraints
1.6 Standards
1.7 Approved Objectives
1.8 Methodology
1.9 Project Outcomes and Deliverables
1.10 Novelty of Work
2. Requirement Analysis 2.1
Literature Survey
2.1.1 Theory Associated with Problem Area
2.1.2 Existing Systems and Solutions
2.1.3 Research Findings for Existing Literature
2.1.4 Problem Identified
2.1.5 Survey of Tools and Technologies Used
2.2 Software Requirement Specification
2.2.1 Introduction
2.2.1.1 Purpose
2.2.1.2 Intended Audience and Reading Suggestions
2.2.1.3 Project Scope
2.2.2 Overall Description
2.2.2.1 Product Perspective
2.2.2.2 Product Features
2.2.3 External Interface Requirements
2.2.3.1 User Interface
2.2.3.2 Hardware Interfaces
2.2.3.3 Software Interfaces
2.2.4 Other Non-functional Requirements
2.2.4.1 Performance Requirements
2.2.4.2 Safety Requirements
2.2.4.3 Security Requirements
2.3 Cost Analysis
2.4 Risk Analysis

5
3. Methodology Adopted
3.1 Investigative Techniques
3.2 Proposed Solution
3.3 Work Breakdown Structure
3.4 Tools and Technology
4. Design Specifications
4.1 System Architecture
4.2 Design Level Diagrams
4.3 User Interface Diagrams
4.4 Snapshots of Working Prototype
5. Conclusions and Future Scope
5.1 Work Accomplished
5.2 Conclusions
5.3 Environmental Benefits
5.4 Future Work Plan
APPENDIX A: References
APPENDIX B: Plagiarism Report

6
 LIST OF TABLES

Table No. Caption Page No.

Table 1 Research Findings for Existing Literature 21-22
Table 2 Cost Analysis 25
Table 3 Investigation techniques 26-27

7
 LIST OF FIGURES

Figure No. Caption Page No.

Figure 1 Overview of our System 17
Figure 2 High level overview of our System 18
Figure 3 Block Diagram for Smart Scan 33
Figure 4 Component Diagram for Smart Scan 34
Figure 5 DFD (Level 0) for Smart Scan 35
Figure 6 DFD (Level 1) for Smart Scan 36
Figure 7 Class Diagram for Smart Scan 37
Figure 8 Use Case Diagram for Smart Scan 39
Figure 9 Activity Diagram for Smart Scan 43
Figure 10 Homepage of Website 45
Figure 11 Marks Page of Website 46

8
 LIST OF ABBREVATIONS

ABBR1 Abbreviation 1
ABBR2 Abbreviation 2

9
 INTRODUCTION

1.1 Project Overview

The imminent introduction of the Automated Grading and Secure Transcript Management System,
"SmartScan," represents a pivotal moment in the evolution of academic assessment. This groundbreaking
system is poised to revolutionize the conventional grading process by harnessing the power of automated
scanning technology, facilitating the seamless transfer of individual question marks to a centralized database.
The inherent elimination of manual data entry not only anticipates a substantial reduction in errors but also
foresees the expeditious progression of the grading process, marking a new epoch of efficiency in academic
evaluation. In Indian education, particularly within centralized examination bodies like the Central Board of
Secondary Education (CBSE), where the sheer volume of answer sheet necessitates a more streamlined
approach, SmartScan emerges as a timely solution to meet the pressing needs of efficiency and accuracy.

SmartScan distinguishes itself with a robust security infrastructure that prioritizes the confidentiality and
integrity of academic records. This commitment to security becomes increasingly crucial within the context
of major examination bodies like CBSE, where the protection of student information is paramount.
Noteworthy is the fact that SmartScan diverges from traditional approaches, eschewing the development of a
graphical user interface. Instead, educators will be presented with a user-friendly Excel sheet, ensuring facile
access to review and validate recorded grades. It provides a simple and accessible method for educators to
navigate and manage academic assessments.

Moreover, SmartScan's automatic sheet unwinding mechanism minimizes manual effort associated with
handling answer sheets, saving valuable time for educators, and enhancing grading efficiency. This
automation, particularly impactful for large entities, is coupled with a simplified Excel output, making
SmartScan a forward-thinking solution for evolving assessment needs in the Indian educational landscape. Its
user-focused approach positions it as an instrumental tool for streamlining processes, enhancing accuracy, and
embracing the challenges of scale within India's diverse educational ecosystem.

10
1.2 Need Analysis

To comprehensively address challenges in the manual grading and mark entry process at Thapar Institute of
Engineering, a survey was conducted among key participants to gauge their perspectives on implementing
Smart Scan.

Identified Needs and Survey Insights:

• Challenges in Manual Process: Participants expressed frustration with the time-consuming manual
grading process, citing difficulties in managing a large volume of answer sheets efficiently.
• Perception on Automated System: Participants viewed Smart Scan as a promising solution to
streamline the grading and mark entry process, acknowledging its potential to alleviate manual
workload.
• Willingness to Adopt Automated System: Participants showed a keen interest in adopting Smart
Scan, recognizing the benefits of increased efficiency and reduced errors in assessment management.
• Efficiency Improvement: Addressing the inefficiencies in the manual grading process, participants
emphasized the need for a solution that would expedite the assessment workflow.
• Error Reduction: Participants highlighted the importance of minimizing errors and inconsistencies in
marks entry, considering the potential impact on students’ academic records.
• Time Savings: Time savings emerged as a crucial factor, with participants expressing the desire to
allocate more time to teaching and academic engagement rather than administrative tasks.
• Enhanced Accessibility: Participants emphasized the need for an accessible and user-friendly
platform that would make grading and mark entry more convenient and efficient.
• Consistent Data Management: Establishing a centralized and automated system for consistent data
management was deemed essential to ensure uniformity in the assessment process.

11
1.3 Research Gaps

12
1.4 Problem Definition and Scope

Grading and managing the examination sheets is an essential task in institutions, which sometimes becomes
tiresome, especially when the number of physical answer sheets is high. Obviously, the old-fashioned approach
of manually keying the marks into a system is very time-consuming and full of human errors—misreading
handwriting, wrong data entry, or even at times the loss of papers. Such errors can lead to huge variations in
the students' records and increased workload for teachers, who double-check them for errors and correct data.
With increasing importance of the standards of education and integrity of data in today's time, comes equally
high demand for an efficient, accurate and automated solution to solve the above-mentioned challenges and
thereby optimize the whole process of examination and grading.

Smart Scan seeks to solve these challenges by automating the entire process of mark detection from
captured images of physical answer sheets. Such a system shall be powered by some major advanced
technologies targeted at an ESP32 CAM module for real-time image capture, a server with pipeline image
preprocessing, and a machine learning model specifically trained for detecting and interpreting marks tables
in answer sheet scans. An OCR engine will then be used to recognize the marks accurately and further
translate them into digital data. The envisioned system will be able to handle hundreds of answer sheets
efficiently, reducing the possibility of human error to a minimum while increasing the grading process
multifold.

13
1.5 Assumptions and Constraints

• Legibility of Marks: The marks on the answer sheets are well readable and written in uniformity. This
will ensure that the OCR engine identifies them correctly.
• Standardised Answer Sheets: All answer sheets use a common template with standardized placing of
mark tables. This makes it easier for detection by the ML model.
• Faultless Configuration of the System: The system, both hardware and software, will be considered
technically correct and error-free.
• Competence of User: The user of this system— instructors—shall have sufficient technical knowledge
about the operation of the system, especially when verifying marks.
• Database Availability: This is a rather simple assumption—that the central database will be available
and responsive all the time—so that the registered marks can be uploaded successfully and without
interruption.
• Legibility and Standardization of Marks: The marks must be clearly legible and written in a standard
format; otherwise, deviations may lead to recognition errors at the hands of the OCR engine.
• Answer Sheet Format Fixed: This system is designed to work with answer sheets only in a predefined
format. Any deformities in the format may lead to different types of detection and recognition errors.
• Hardware and Software Capabilities: The performance of the entire system, totally dependent on
hardware and software capabilities, is limited.
• Human Verification Needed: The marks, recognized automatically by the system, need human
verification before finally submitting the results. This adds an additional time factor to the whole
process.
• Language and Symbol Recognition: The OCR engine is designed only for the recognition of marks
in certain languages/symbols only. The usage of unsupported languages or symbols may lead to
misrecognition.

14
1.6 Standards

To ensure reliability and usability in Smart Scan, the following key standards are essential:

1. Data Standards
- Format: Use CSV or JSON for data storage to ensure compatibility.
- Integrity: Apply checksum algorithms to maintain data accuracy during transfers.
2. OCR and Image Processing Standards
- OCR Accuracy: Achieve at least 99% accuracy for reliable mark recognition.
- Resolution: Ensure a minimum of 720p for clear image capture by the ESP32 CAM.
3. User Interface and Accessibility Standards
- Design: Follow Material Design for a user-friendly interface.
- Accessibility: Comply with WCAG 2.1 to make the system accessible to all users.
4. Security Standards
- Privacy: Adhere to GDPR for data protection.
- Authentication: Implement Email login for secure user access.
5. Software Development Standards
- Coding: Use consistent coding practices like PEP 8 for maintainability.
- Version Control: Employ Git with semantic versioning for effective project management.

15
1.7 Approved Objectives

1. Efficiency Improvement in Information Capture: Enhance the efficiency of information capture from
the first page of the answer sheet, ensuring quick and accurate extraction of relevant details.
2. Error Reduction in Information Extraction: Develop algorithms and mechanisms within the system that
significantly reduce errors in extracting information from the first page, ensuring precise and reliable data.
3. Time Savings in Administrative Tasks: Design the system to save administrative time by automating the
process of capturing information from the first page, allowing educators to focus more on teaching and
academic engagement.
4. Enhanced Accessibility for Information Retrieval: Create a user-friendly and accessible platform that
simplifies the process of retrieving information captured from the first page, ensuring educators can easily
access and utilize the data.
5. Consistent Data Management for Captured Information: Establish a centralized and automated data
management system that ensures consistency in handling information captured from the first page,
promoting standardized processes.

16
1.8 Methodology

Step 1: Pre-processing to Find Coordinates (Answer Sheet Preprocessing)

1. Objective:
o Enhance images and identify predefined boxes to establish coordinates for mark extraction.
2. Tasks:
o Implement image enhancement using OpenCV to improve clarity and reduce noise.
o Utilize edge detection and contour finding techniques to identify predefined boxes on the answer
sheet.
o Determine precise coordinates of the identified boxes for subsequent mark extraction.

Step 2: Software Development (Including ICR)

1. Objective:
o Develop software for image capture, pre-processing, Intelligent Character Recognition (ICR) for
mark and roll number extraction, and database integration.
2. Tasks:
o Create modules for image capture with ESP32-CAM and Wi-Fi transmission.
o Implement image pre-processing algorithms for enhancing captured images.
o Develop ICR algorithms using OpenCV to intelligently recognize and extract marks and roll
numbers.
o Integrate a secure database system for storing and managing extracted marks and associated roll
numbers.
o Implement a user-friendly interface for educators.

Figure 1 Overview of our System

Step 3: Hardware Development using Roller and Arduino

1. Objective:
o Design and develop the hardware components for automating the sheet unwinding process.

17
 2. Tasks:
o Integrate a roller mechanism controlled by an Arduino for automated sheet unwinding.
o Establish communication between the computer and Arduino to trigger the unwinding process
based on the success signal.
o Ensure synchronized operation between the hardware and software components.

Figure 2 High level overview of our System

Step 4: Integration and Testing
1. Objective:
o Integrate the software and hardware components for a cohesive system.
o Conduct thorough testing to ensure functionality, accuracy, and reliability.
2. Tasks:
o Integrate the software modules with the hardware components.
o Conduct initial tests to validate image capture, pre-processing, ICR, and mark extraction.
o Verify the automated unwinding mechanism’s accuracy and responsiveness.
o Perform system-wide testing to ensure seamless operation.

Step 5: User Training and Feedback

1. Objective:
o Provide training to educators on system usage.
o Gather user feedback for further refinements.
2. Tasks:
o Conduct training sessions for educators on utilizing the Automated Top Sheet Mark Capturing
System.
o Collect feedback to identify any user experience improvements.
o Implement necessary adjustments based on user feedback.

18
This comprehensive methodology outlines a step-by-step approach, incorporating pre-processing, software
development with ICR, hardware integration, testing, and user training. The timeline provides a structured
plan for the development of Smart Scan System.

1.9 Project Outcomes and Deliverables

The outcome of the Smart Scan project will be a sophisticated software solution capable of accurately
recognizing and recording marks from handwritten answer sheets. The system will offer the following key
features and benefits:

• Accurate Mark Recognition: Utilizing advanced image processing algorithms and machine learning
techniques, computer vision in this case, the system will accurately identify, and extract marks written
on answer sheets, ensuring reliable grading results.

• Efficient Processing: By automating the marking process, the system will significantly reduce the
time and effort required for grading examinations, allowing educators to focus on more value-added
tasks.

• User-Friendly Interface: The system will feature an intuitive user interface(hardware) that enables
educators to easily upload answer sheets, initiate the marking process, and retrieve graded results
conveniently.

• Reliability and Consistency: With a focus on robustness and accuracy, the system will consistently
produce reliable grading outcomes, minimizing the potential for human error and ensuring fairness in
assessment.

• Scalability: Designed to handle large volumes of answer sheets, the system will be scalable to
accommodate varying examination sizes and formats, making it suitable for educational institutions of
all sizes.

• Time Savings: By automating the marking process, the system will save considerable time for
educators, allowing them to allocate their time and resources more efficiently towards other aspects of
teaching and learning.

Customization and Adaptability: The system will offer flexibility for customization and adaptation to
different educational contexts, including support for multiple languages and examination types.

19
1.10 Novelty of Work

The novelty of Smart Scan lies not in the core concept of automated grading, which exists in various forms,
but in its specific design choices and implementation:

1. Streamlined data transfer: Instead of individual answer sheets feeding into a complex software
system, Smart Scan transfers individual question marks directly to a centralized database, simplifying
the process and potentially reducing errors.
2. Focus on large-scale efficiency: Smart Scan is specifically designed for high-volume assessments
like those conducted by central examination bodies in India, catering to the need for efficient and
accurate grading of massive amounts of data.
3. User-friendly interface: Instead of relying solely on Excel sheets, Smart Scan will offer a user-
friendly graphical user interface (UI) specifically designed for educators' needs. This UI will provide
intuitive navigation, clear data visualization, and easy access to functionalities, making the review and
validation of recorded grades efficient and user-friendly.
4. Automatic sheet unwinding: This feature reduces manual effort and saves time, especially when
dealing with large numbers of answer sheets.

In essence, Smart Scan's novelty lies in combining existing technologies (automated grading) with specific
design choices (data transfer, security, user interface, automation) to address the needs of large-scale
assessment within the Indian educational context.

20
 REQUIREMENT ANALYSIS

2.1 Literature Survey

2.1.1 Theory associated with problem area

The area of concern is automating the handwritten mark recognition and scanning process used to
evaluate answer sheets. Techniques for image processing and optical character recognition (OCR) are
crucial for the same. OCR is a technology that turns a variety of document types - such as PDFs,
scanned paper documents, and digital camera images -into editable and searchable data. OCR is used
to extract numerical data from handwritten answer sheets to scan each student's mark sheet.

2.1.2 Existing problem and solutions

There are several programs out there that try to automate result management and grading, including:
• Optical mark recognition (OMR) systems are frequently employed in standardized assessments.
OMR systems record responses by scanning checkboxes or bubbles on forms. Nevertheless, they
cannot handle free-form numerical grading and are restricted to pre-defined answer sheets.
• Machine Learning-Based Evaluation: While some solutions exist that use machine learning to
grade subjective responses, they are still in the early stages of development and frequently need
large amounts of training data to function well.

2.1.3 Research Findings for Existing Literature

Title Authors Year Key Findings

HCR-Net: A deep learning- Vinod Kumar 2024 Introduced HCR-Net, a novel script-
based script independent Chauhan, Sukhdeep independent deep learning network for
handwritten character Singh & Anuj Sharma Handwritten Character Recognition
recognition network (HCR). Achieved significant
improvements across various languages,
set 26 new benchmarks, and reduced
trainable parameters. Publicly released
code for reproducibility.

OECA-Net: A co-attention Feng Yan, Wushouer 2023 Developed a co-attention model for Visual
network for visual question Silamu, Yachuang Question Answering (VQA) that enhances
answering based on OCR Chai & Yanbing Li text understanding in images using OCR
scene text feature token integration and question-guided
enhancement attention, leading to improved

21
 performance in scene text-related
questions.

English Handwritten Shrinivas R. Zanwar, 2023 Proposed EfficientDet-D4 for Handwritten

Character Recognition Yogesh H. Bhosale, Digit Recognition (HDR), addressing
Based on Ensembled Devendra L. Bhuyar, challenges like diverse writing styles and
Machine Learning Zakee Ahmed, Ulhas image artifacts. Achieved 99.83%
B. Shinde & accuracy on MNIST and 99.10% on USPS
Sandipann P. Narote datasets.

A Novel Technique for Syed Soahil Ahmed, 2023 Introduced a deep learning approach using
Handwritten Digit Zahid Mehmood, EfficientDet-D4 for HDR, overcoming
Recognition Using Deep Imran Ahmad Awan & challenges like writing style diversity and
Learning Rehan Mehmood image artifacts, with 99.83% accuracy on
Yousaf MNIST and 99.10% on USPS datasets.

Neural representations of Juhyeon Lee & 2023 Investigated neural representations for
the perception of Minyoung Jung visual perception of handwritten digits
handwritten digits and using fMRI and VGG16-based CNN.
visual objects from a Found hierarchical alignment between
convolutional neural CNN features and human visual areas,
network compared to highlighting unique human neural
humans representations.

MNIST Handwritten Digit Haijian Shao, Edwin 2022 Presented a deep CNN model for HDR
Classification Based on Ma, Ming Zhu, Xing with optimized hyperparameters,
Convolutional Neural Deng & Shengjie Zhai achieving 99.82% accuracy on training
Network with data and 99.40% on testing data for
Hyperparameter MNIST. Highlighted fast convergence and
Optimization high effectiveness.

Enhanced handwritten digit Debabrata Swain, 2023 Addressed handwritten digit recognition
recognition using optimally Badal Parmar, Hansal through an ANN optimized with Adagrad,
selected optimizer for an Shah, Aditya Gandhi, achieving 98.55% accuracy, critical for
ANN Biswaranjan Acharya document verification and mail
& Yu-Chen Hu processing.

Table 1: Research Findings for Existing Literature

22
 2.1.4 Problems Identified
The main hurdle is making sure the system can handle different degrees of handwriting styles and
reliably recognize and document handwritten marks from a variety of handwriting styles. Furthermore,
there are several important considerations, including how to incorporate this functionality into a facile
web interface and guarantee data security throughout the upload and storage procedures. In addition, a
key concern is the software's ability to integrate with the hardware, which includes rollers and scanners.

2.1.5 Survey of Tools and Technology Used

Smart Scan is built using React for creating a responsive and dynamic frontend, paired with Firebase
for managing real-time data, authentication, and storage. Python is the core language, integrating
essential libraries like YOLO for real-time object detection and OpenCV for advanced image
processing tasks.

To streamline the workflow, Roboflow is utilized for data annotation and model training, simplifying
the process of working with image datasets. This setup allows for efficient implementation of machine
learning models that can accurately identify and capture marks from scanned documents.

The system is designed for scalability and reliability, leveraging cloud platforms like Google Cloud to
ensure seamless data management and deployment. Internally Firebase enhances the system's ability
to operate smoothly across different environments, providing a solid foundation for accurate marks
recognition.

2.2 Software Requirement Specification

2.2.1 Introduction
2.2.1.1 Purpose
The intended purpose of this project is to create an automated system that will reduce the workload
of our teachers by scanning, extracting marks from answer sheets, and uploading these scores to a
central website for further analysis and record-keeping.

2.2.1.2 Intended Audience and Reading Suggestions

The developers, faculty members, and panel heads involved in this project are the intended audience
for this document. It ought to operate as a guide during the development process as it describes the
specifications and goals for the project.

2.2.1.3 Project Scope

The project's scope encompasses the creation of an online application that enables educators to use
OCR technology to extract marks from answer sheets and upload the results to a secure online
platform such as the webkiosk. The automated system should be able to recognize handwritten
marks accurately and handle different kinds of answer sheets.

2.2.2 Overall Description

23
 2.2.2.1 Product Perspective
The "Smart Scan" project is a web application that can be integrated with Webkisok or LMS, both
widespread educational management systems. The objective is to optimize the grading procedure
and minimize manual labour by means of automating the extraction and storage of marks.

2.2.2.2 Product Features

• OCR Integration: Using scanned answer sheets, automatically identify and extract marks.
• User-Friendly Interface: An easy-to-use interface for viewing results and uploading answer
sheets.
• Data Storage: Capable of accessing and analyzing data, securely storing scanned marks.
• Error Handling: To minimize confusion, any errors in the marks uploaded can be manually
resolved.

2.2.3 External Feature Requirements

2.2.3.1 User Interfaces
The application will offer a web-based user interface that can be reached via common web browsers.
The user interface will have tools for managing data, viewing scanned results, and displaying answer
sheets.

2.2.3.2 Hardware Interfaces

Standard computer accessories, like a camera for grabbing answer sheets, will be needed for the
system. Rollers would also be necessary to guarantee the answer sheets moved without difficulty.

2.2.3.3 Software Interfaces

To identify the marks, the system will interface with OCR libraries. React would be used for the
front end and Node for the back end of the website's creation. Also, it will establish a connection
with a backend database for storing the results and acquired data.

2.2.4 Other Non-Functional Requirements

2.2.4.1 Performance Requirements
In a matter of seconds, the system ought to be able to process an answer sheet and extract the marks.
Additionally, it needs to be competent to manage several concurrent uploads without noticeably
deteriorating performance.

2.2.4.2 Safety Requirements

To safeguard against loss, the system needs to make sure that the data is frequently backed up.
Furthermore, security measures against unauthorized access and tampering of the scanned results
must be in place.
24
 2.2.4.3 Security Requirements
Data transmission through the system must be carried out via secure communication channels. To
assure that only individuals with permission can access sensitive data, user authentication and
authorization policies should be strictly followed.

2.3 Cost Analysis

Item Name Estimated Cost (Rs.)

ESP 32 Camera 2500

Structure(Framework) Materials and Making 3000

Batteries and Wires 1500

Total (Estimated Cost) 7000

Table 2: Cost Analysis

2.4 Risk Analysis

• OCR Accuracy: There's a chance that not all handwriting styles will be correctly recognized by OCR.
• Data Security: The possibility of data leaks when storing or uploading data.
• Scalability: The possibility that the system won't function properly when under a lot of stress,
especially during busy times like exam periods.
• Hardware Integration: The risk is minimizing user effort and ensuring a hassle-free movement of the
answer sheets by carefully setting up the rollers to prevent external chaos.

25
 METHODOLOGY ADOPTED

3.1 Investigation Techniques

S. No. Investigate Investigative Description Techniques Investigative Project

Project Examples
Techniques

1 Literature Literature review was conducted at the very Reviewing the available
Review initial stage of the investigation. It involved research papers, articles, and
an in-depth survey of the prevailing case studies to identify
research and developments in the sphere of algorithms, already existing
a Smart Scan with a special focus on frameworks, and systems
computer vision, deep learning, and real- that are relevant to Smart
time data processing. The main objective Scan.
was to identify best practices and new
techniques to use in the Smart Parking
System. The review was supposed to be
useful in attaining a number of insights and
innovations useful during the design and
implementation stages of the system
through scrutiny of previous works.

2 Case Study This case study analysis technique Analyzing case studies of
Analysis involved examining the already existing character recognition
Smart Scan that are deployed in various systems that use OpenCV
cities. The analysis is focused on for computer vision and
understanding challenges, learning from deep learning models for
the success stories of the existing systems marks detection
and studying how different technologies
were integrated together.

3 Technical The technical feasibility study was Testing Python libraries like
Feasibility supposed to answer the question of whether OpenCV and TensorFlow,
Study the technologies chosen for integration into evaluating different cameras

26
 Smart Scan are viable in terms of their and processors, and
practicality. That means the testing of simulating large-scale
several libraries and frameworks for deployments to ensure
appropriateness, looking at different system scalability and
hardware configurations with regard to reliability.
compatibility and performance, and
studying the system's scalability to prove
the functionality under conditions of future
growth. Besides, environmental factors are
taken into consideration to verify that all
the subsystems work perfectly under real-
world circumstances. It was conducted
with a thorough assessment to ensure that
the technologies chosen for this purpose
would support such goals of the system in
an efficient and reliable way.

4 Prototyping Prototyping and simulation were used to Developing prototypes of

and test and refine the system in a controlled the computer vision module,
Simulation environment before full-scale deployment. simulating traffic levels, and
These techniques helped in early detection testing user interface designs
of issues, iterative improvement, to ensure they are intuitive
performance benchmarking, and user and user-friendly.
interface testing.

Table 3: Investigation Techniques

27
3.2 Proposed Solution

To address the inefficiencies of manual mark management, Smart Scan is designed as a web-based application
that provides a secure, efficient and user-friendly platform to students, teachers and administrators. The system
will use modern web technologies and cloud services to ensure scalability, reliability and ease of use. The
front-end of the application will be developed using React, a robust JavaScript library known for creating
responsive and interactive user interfaces. The backend will be powered by Firebase, which offers a
comprehensive cloud infrastructure, including Firebase Authentication for secure login and identity
verification and Firebase Firestore for real-time data storage and management.

The system will include different interfaces tailored to the needs of students, teachers and administrators. The
student interface will allow users to log in securely using Firebase Authentication and access a personalized
dashboard where they can view their grades for different subjects. The dashboard will display tags in a clear
and organized manner with data fetched in real-time from Firebase Firestore. This ensures that students always
have up-to-date information about their study results.

For teachers, the system will provide a secure login process and a dedicated interface for managing student
evaluations. Teachers will be able to update grades for individual students through a user-friendly interface
with data stored directly in Firebase Firestore. Additionally, teachers or administrators will have the ability to
review and approve updated assessments. Once approved, labels are locked in the system, preventing further
modifications and ensuring data integrity.

Administrators will have a comprehensive interface that allows them to review and approve all student
assessments. This final approval step will act as a quality check to ensure the labels are accurate before they
are locked. The system's data security will be further enhanced by role-based access control, which ensures
that only authorized users have access to specific functions. By locking tags after approval, the system
maintains the integrity and accuracy of stored data and prevents unauthorized changes.

Smart Scan is designed to be scalable and able to handle a large number of users simultaneously without
performance degradation. Firebase's cloud infrastructure will support the scalability needs of the system and
ensure that it is responsive and efficient as the user base grows. Overall, this solution will streamline the
student grade management process, reduce the likelihood of errors, and increase the overall efficiency of the
academic institution's grading system. Thanks to the use of React and Firebase, the system will be easy to
maintain and will be able to meet the requirements of modern educational environments.

28
3.3 Work Breakdown Structure (WBS)

During the development of Smart Scan, several key modules or products can be identified that present a
critical component of the solution in its entirety. The WBS produces a clear roadmap for the project and
explains all tasks with deliverables associated to ensure that the objectives of the project are realized.

1. System Design and Planning

• 1.1 Define System Architecture
o Design overall system architecture (frontend, backend, database)
o Map interactions between components (students, teachers, administrators)
• 1.2 Identify Hardware and Software Requirements
o Select technology stack (React, Firebase, etc.)
o Determine any additional tools or resources needed
• 1.3 Develop Project Timeline and Milestones
o Create a project timeline with key milestones
o Define deliverables and deadlines

2. Data Collection and Preprocessing

• 2.1 Setup Authentication and Database
o Configure Firebase Authentication for user management
o Set up Firebase Firestore for data storage
• 2.2 Collect Initial Data
o Gather sample student marks and other necessary data for initial testing
• 2.3 Preprocess Data
o Clean and normalize data to ensure consistency
o Implement data validation rules for integration with the system

3. Development of Core Functionalities

• 3.1 Develop Student Interface
o Implement login and authentication for students
o Create a dashboard for students to view their marks
• 3.2 Develop Teacher Interface
o Implement login and authentication for teachers
o Develop features for updating and approving marks
• 3.3 Develop Administrator Interface
o Implement login and authentication for administrators
o Create tools for reviewing, approving marks, and generating reports

29
4. Real-Time Data Processing and Analysis
• 4.1 Design and Implement Data Processing Logic
o Develop scripts for handling real-time data updates and interactions
• 4.2 Develop Real-Time Data Communication Protocols
o Ensure efficient communication between frontend and backend
• 4.3 Optimize System Performance
o Conduct performance testing and optimization for real-time analysis
o Address any performance bottlenecks

5. User Interface Development

• 5.1 Design and Implement Student Interface
o Develop the student dashboard to display marks
o Ensure a user-friendly and accessible design
• 5.2 Develop Teacher Interface
o Implement features for teachers to manage and approve marks
• 5.3 Develop Administrator Interface
o Implement tools for administrators to review and finalize marks
o Create functionalities for generating reports

6. System Integration and Testing

• 6.1 Integrate System Components
o Combine frontend, backend, and database components
• 6.2 Conduct System Testing
o Perform functional testing for each user role (student, teacher, administrator)
o Conduct security and performance testing
• 6.3 Perform User Acceptance Testing (UAT)
o Test with end-users and gather feedback
o Make adjustments based on feedback

7. Deployment and Monitoring

• 7.1 Deploy the System
o Set up Firebase Hosting and deploy the application
o Configure domain and SSL certificates for secure access
• 7.2 Set Up Monitoring and Maintenance Protocols
o Implement monitoring tools to track system performance
o Plan for ongoing maintenance and updates
• 7.3 Train End Users
o Conduct training sessions for students, teachers, and administrators
o Provide user manuals and support materials

30
8. Post-Deployment Support and Continuous Improvement
• 8.1 Provide Ongoing Technical Support
o Offer support for technical issues and user queries
• 8.2 Monitor System Performance
o Regularly review system performance metrics and make updates as needed
• 8.3 Collect and Analyse User Feedback
o Gather feedback from users to identify areas for improvement
o Implement enhancements based on user feedback

Each of these modules is crucial for the successful development and deployment of the Automated Marks
Capturing System. The WBS helps in breaking down the project into manageable parts, ensuring that all
components are developed systematically and efficiently, leading to a well-coordinated final product.

3.4 Tools and Technology

Smart Scan will be built using a combination of modern tools and technologies to ensure a robust, scalable
and user-friendly application. From frontend development to backend services and database management,
these tools and technologies have been carefully selected to meet the specific needs of the system.

Frontend Development will be handled using React, a popular JavaScript library known for its efficiency in
creating dynamic and responsive user interfaces. React allows developers to create reusable components,
which not only speeds up the development process, but also ensures a consistent user experience across the
application. With its virtual DOM, React optimizes rendering, making the application faster and more
responsive. This is particularly important in an educational environment where multiple users may be
accessing the system simultaneously.

On the Backend, Firebase will serve as the backbone of the system. Firebase offers a comprehensive set of
tools that are necessary for the functionality of the application. Firebase Authentication will be used to log
in and authenticate users, providing secure and reliable identity verification for students, teachers, and
administrators. The service supports various authentication methods, including email/password providers and
OAuth, making it flexible for different user requirements.

Firebase Firestore will be the primary database used for data storage and management. Firestore is a cloud-
based NoSQL database that enables real-time data synchronization, which is essential for an application where
students and teachers need to access and update grades simultaneously. Its ability to automatically scale based
on application needs ensures that the system can handle growing user numbers without degrading performance.
In addition, Firestore security features such as role-based access control and data validation rules will be used
to protect sensitive student information and ensure data integrity.

31
For Hosting and Deployment, Firebase Hosting is used to deploy and serve the web application. Firebase
Hosting is a fast and secure web hosting service that seamlessly integrates with other Firebase services to
provide a comprehensive solution for application development and deployment. It supports global content
delivery and ensures that users have fast and reliable access to the system no matter where they are.

To make the development process efficient and the application maintainable, Git will be used for version
control. Git allows multiple developers to work on a project simultaneously, track changes, and collaborate
effectively. This version control system is essential for managing the development lifecycle, especially in a
project that may involve constant updates and improvements.

In short, Smart Scan will use a combination of React for frontend, Firebase for backend services and database
management, and Git for version control. These tools and technologies are suitable for building a secure,
scalable and efficient system that meets the needs of students, teachers and administrators. By utilizing these
modern technologies, the project will not only meet current requirements, but also remain adaptable to future
needs and improvements.

32
 DESIGN SPECIFICATIONS

4.1 System Architecture

4.1.1 Block Diagram

Figure 3: Block Diagram for Smart Scan

Explanation of the Above Diagram:

➢ Input Module:
▪ Image Acquisition: This component is responsible for capturing images of student marksheets or
answer sheets.
▪ Wi-Fi Module: The captured images are transmitted to the central system through a Wi-Fi
connection, facilitating the data transfer process.

➢ Recognition Module:
▪ Preprocessing: The acquired images undergo preprocessing to prepare the data for accurate
recognition.

33
 ▪ Machine Learning Model: A machine learning model is employed to analyze the pre-processed
data, extracting relevant information such as marks.
▪ Optical Character Recognition (OCR): OCR technology is used to convert the extracted data from
images into text format, identifying characters like student names, roll numbers, and marks.

➢ Output Module:
▪ Validation by User: After the recognition process, the extracted data is presented to teachers or
administrators for validation, ensuring accuracy before finalization.
▪ Results Storage: Once validated, the finalized marks are securely stored in the database (e.g.,
Firebase Firestore) for future retrieval and reporting.

➢ User Interface:
▪ Dashboard: Provides an interface for students, teachers, and administrators to interact with the
system, including viewing, updating, or approving marks.
▪ Generate Analytics: Allows administrators to generate reports and analytics based on the stored
data, such as class averages and individual student performance.

4.1.2 Component Diagram

Figure 4: Component Diagram for Smart Scan

Explanation of the Above Diagram:

❖ Input Module:
▪ Image Acquisition: This component is responsible for capturing images of student marksheets or
answer sheets.
▪ Wi-Fi Module: The captured images are transmitted to the central system through a Wi-Fi
connection, facilitating the data transfer process.

34
 ❖ Recognition Module:
▪ Preprocessing: The acquired images undergo preprocessing to prepare the data for accurate
recognition.
▪ Machine Learning Model: A machine learning model is employed to analyze the pre-processed
data, extracting relevant information such as marks.
▪ Optical Character Recognition (OCR): OCR technology is used to convert the extracted data
from images into text format, identifying characters like student names, roll numbers, and marks.

❖ Output Module:
▪ Validation by User: After the recognition process, the extracted data is presented to teachers or
administrators for validation, ensuring accuracy before finalization.
▪ Results Storage: Once validated, the finalized marks are securely stored in the database (e.g.,
Firebase Firestore) for future retrieval and reporting.

❖ User Interface:
▪ Dashboard: Provides an interface for students, teachers, and administrators to interact with the
system, including viewing, updating, or approving marks.
▪ Generate Analytics: Allows administrators to generate reports and analytics based on the stored
data, such as class averages and individual student performance.

4.2 Design Level Diagrams

4.2.1 Data Flow Diagrams (DFD)

Level 0:

Figure 5: DFD (Level 0) for Smart Scan

35
 Level 1:

Figure 6: DFD (Level 1) for Smart Scan

Explanation Of DFD:

Level 0:
1. Smart Scan: The Level-0 Data Flow Diagram (DFD) provides a high-level overview of the system’s
interactions. The main actors in the system are the teacher/instructor and the database/CSV. The teacher
can enter a stack of answer sheets into the system, update/edit marks as necessary, and fetch and view
marks. The system interacts with the database/CSV to store and retrieve data as needed, ensuring the
smooth functioning of Smart Scan.

Level 1:
1. ESP32 CAM Driver:
The Level-1 DFD expands on the process of capturing the answer sheets. The ESP32 CAM Driver is
responsible for capturing images of the answer sheets and sending the feed via RTSP (Real-Time
Streaming Protocol) for further processing. This is the first step in the automated marks recognition
process.

2. Image Pre-processing:
Once the images are captured, they are sent to the image pre-processing module. This step involves
processing the raw images to enhance quality, adjust brightness/contrast, and remove any noise,
ensuring that the images are ready for accurate recognition of the marks.

3. Model to Recognize Marks Table:

The processed images are then fed into a model designed to recognize the marks table on the answer
sheets. This model identifies the coordinates of the marks table, enabling the system to focus on the
relevant sections of the sheet for further analysis.

4. Intelligent Character Recognition (ICR):

After the marks table is identified, the Intelligent Character Recognition (ICR) system is used to scan
and recognize the individual marks. The ICR system translates the scanned marks into digital data that
can be uploaded to the central database.

36
5. Database Drivers:
The scanned and recognized marks are then uploaded to the central database using database drivers.
This step ensures that all marks are securely stored and can be easily accessed or modified as needed.

6. Teacher Interaction:
The teacher can interact with the system by requesting to view or update marks. Any changes made by
the teacher are acknowledged by the system, ensuring that the marks data is accurately reflected in the
database. This interaction loop ensures that the teacher can manage the marks data efficiently while
the system handles the automation of marks recognition.

4.2.2 Class Design Diagram

Figure 7: Class Diagram for Smart Scan

Explanation of the Above Diagram:

❖ AnswerSheet:
The AnswerSheet class represents a student's physical exam paper where multiple marks are recorded.
In a real-life scenario, this is equivalent to collecting all exam sheets after a test. The system stores
these marks digitally, allowing for easy retrieval and management. Each mark on the exam paper is
captured and stored as an individual `Mark` object within the AnswerSheet.

❖ Mark:
The Mark class holds the value of each individual score a student receives on their exam. In real life,
this corresponds to the actual grade a student earns in each subject. For example, if a student scores 90
in science, this score is stored as a `Mark` object. This class allows the system to manage these scores,
making it possible to retrieve, update, or display them as needed.

37
❖ OCR (Intelligent Character Recognition):
The OCR class is responsible for extracting marks from images of the answer sheets. In practice, this
represents the technology that scans and reads the handwritten or printed marks on physical exam
papers. For instance, after scanning an answer sheet, the ICR software identifies the numbers written
as scores and converts them into digital data that the system can process.

❖ ImagePreprocessor:
Before the marks are extracted by the ICR, the ImagePreprocessor enhances the image quality to ensure
accuracy. In real-life, this is similar to adjusting the brightness or contrast of a photo before scanning
it. The preprocessing step improves the clarity of the marks, making it easier for the ICR software to
accurately read and interpret the scores.

❖ Database:
The Database class is where all the extracted marks are stored. In a real-world scenario, this is like the
school’s digital record system where all student grades are saved. This database allows for secure
storage and easy retrieval of marks whenever they are needed, such as for generating report cards or
academic transcripts.

❖ UserInterface:
The UserInterface class represents the digital platform where students, teachers, and administrators
interact with the system. In a real-life context, this could be an online portal or app where students log
in to view their grades, teachers update scores, and administrators generate reports. It is the primary
way users’ access and interact with the stored data, making the system user-friendly and accessible.

38
4.3 User Interface Diagrams

4.3.1 Use Case Diagram

Figure 8: Use Case Diagram for Smart Scan

39
4.3.2 Use Case Templates

Use Case Template to View Marks

Field Content

Use Case Title View Marks

Abbreviated Title View Marks

Use Case ID 1

Actors User

Description This use case describes how a user views their marks in Smart
Scan.

Pre Conditions User has valid login credentials

Task Sequence 1. User navigates to the login page of the marks updating
website.

2. User enters their login credentials.

3. System authenticates the user.

4. User is directed to their dashboard.

5. User selects the option to view marks.

6. System displays the user's marks for all subjects.

Post Conditions User can see their updated marks for all subjects.

Modification History 20-Aug-2024 Author Ishaan Gaba

Alternate Flow If authentication fails, user is prompted to re-enter credentials

or contact support.

40
 Use Case Template to Approve Marks

Field Content

Use Case Title Review and Approve Marks

Abbreviated Title Approve Marks

Use Case ID 2

Actors Teacher, Administrator

Description This use case describes how a teacher or administrator reviews

and approves marks for final submission in Smart Scan.

Pre Conditions Marks have been updated by the teacher and are awaiting
approval.

Task Sequence 1. Teacher or Administrator logs into the system.

2. Teacher or Administrator navigates to the approval section.

3. Teacher or Administrator reviews the updated marks.

4. Teacher or Administrator approves the marks.

5. System locks the marks to prevent further changes.

Post Conditions Marks are approved and locked in for final submission.

Modification History 20-Aug-2024 Author Shrey Pachauri

Alternate Flow If approval is denied, the system prompts the teacher to make
necessary corrections before resubmitting for approval.

41
 Use Case Template to Generate Marks Report

Field Content

Use Case Title Generate Marks Report

Abbreviated Title Marks Report

Use Case ID 3

Actors Teacher, Administrator

Description This use case describes how a teacher or administrator generates

a report of users' marks in Smart Scan.

Pre Conditions Marks have been entered and approved.

Task Sequence 1. Teacher or Administrator logs into the system.

2. Teacher or Administrator navigates to the reports section.

3. Teacher or Administrator selects the criteria for the report

(e.g., by class, by subject).

4. System generates the marks report based on the selected

criteria.

5. Teacher or Administrator downloads or prints the report.

Post Conditions A report of users' marks is generated and can be downloaded or

printed.

Modification History 20-Aug-2024 Author Abhinav Maheshwari.

Alternate Flow If the report generation fails, the system provides an error
message and prompts the user to retry or contact support.

42
 4.3.3 Activity Diagram

Figure 9: Activity Diagram for Smart Scan

Explanation of the above Diagram:

1. Teacher/Instructor:
o Inserts a Bundle of Answer Sheets: The process begins with the teacher or instructor placing a
stack of answer sheets into the system. This action initiates the automated marks recognition
process.
o Marks Verification: Once the marks are processed and recognized by the system, the teacher
verifies the accuracy of the marks before they are uploaded to the database. This is an essential
step to ensure that the recognized marks are correct.

43
2. System:
o Rolls the First Answer Sheet: The system is responsible for moving the topmost answer sheet
from the stack under the camera for scanning. If the system fails to do this, the process cannot
continue, and the teacher may need to intervene.
o Verification Decision: After the marks are recognized, the system prompts the teacher for
verification. If the teacher verifies the marks, they are uploaded to the database; otherwise, the
process may require additional review or reprocessing.

3. ESP32 CAM:
o Scans the Topmost Sheet: The ESP32 CAM captures an image of the topmost answer sheet and
sends this feed via RTSP (Real-Time Streaming Protocol) to the server for processing. This
camera feed is crucial for the subsequent steps in the process.

4. Server:
o Applies Preprocessing Pipeline: Upon receiving the image feed, the server applies a
preprocessing pipeline. This step involves enhancing the image to improve the accuracy of the
marks recognition by the ML model. Preprocessing may include adjusting contrast, removing
noise, and other image enhancements.

5. ML Model:
o Detects the Marks Table: The machine learning model attempts to detect the marks table on the
answer sheet. If it successfully detects the table, the process continues; otherwise, the server
retries reading the frame to attempt detection again.
o Retries if Necessary: If the marks table is not detected, the server will retry the process by
reprocessing the frame. This ensures multiple attempts are made to accurately locate the table
before proceeding.

6. OCR Engine:
o Attempts to Read Marks: Once the marks table is detected, the OCR (Optical Character
Recognition) engine attempts to read the marks from the table. If the OCR engine reads the
marks with sufficient accuracy, it proceeds to the next step.
o Handles Read Failures: If the OCR engine fails to read the marks correctly, it may either retry
or flag the issue, depending on the system's configuration. Successful reads are then forwarded
for database entry.

7. Database:
o Makes Appropriate Entry: Upon successful recognition of the marks, the system makes an
appropriate entry in the marks table within the database. This step ensures that the recognized
marks are stored securely and can be accessed later for analysis or reporting.

44
 This activity diagram illustrates the flow of tasks across different components (Teacher/Instructor,
System, ESP32 CAM, Server, ML Model, OCR Engine, and Database) in the automated marks
recognition process. Each component has its specific responsibilities, and the diagram highlights the
sequential flow of data and decision-making processes across these components.

4.3.4 Diagram of Prototype

Figure 10: Homepage of Website

45
Figure 11: Marks Page of Website

46
 CONCLUSION AND FUTURE SCOPE

5.1 Work Accomplished

❖ Project Planning & Requirement Gathering:

➢ Outlined the main goals and parameters of the project, as well as important components like interfaces
for teachers, administrators, and students.
➢ Gathered and recorded stakeholder requirements, detailing the features required for every user role.

❖ System Design & Architecture:

➢ React was used for the front end and Firebase was used for the back end and authentication services
when designing the system architecture.
➢ Developed data models in Firebase Firestore to manage user profiles, approval workflows, and marks.
➢ For the dashboards of teachers, administrators, and students, wireframes and mockups were created.

❖ Front-End Development:
➢ Configure the dependencies and project structure, as well as set up the React project environment.
➢ Firebase Authentication was used in the development of the login and authentication interfaces to
guarantee safe access for administrators, teachers, and students.
➢ The student dashboard was put into place, enabling students to see the marks they had pulled from
Firebase Firestore.
➢ Developed the teacher interface, which includes validation to guarantee data accuracy, and allows for
managing and updating student marks.
➢ Created the administrator interface to ensure that only approved marks are finalized by reviewing,
approving, and generating reports.

❖ Back-End Integration:
➢ Role-based access control and safe login are handled by integrated Firebase Authentication.
set up Firebase Firestore to serve as the primary database for markings data retrieval and storage.
➢ CRUD operations, such as safe data writing and reading from Firestore, were implemented for mark
management.
➢ To protect data, set up Firebase rules that limit access according to user roles (administrators, teachers,
and students).

❖ Implementing the Workflow for Approval:

➢ Created the mark approval procedure, wherein educators can send in their marks for examination and
administrators can approve them.

47
 ➢ Added the ability to lock marks after they have been approved, stopping future changes and
guaranteeing data integrity.

❖ Report Creation:
➢ Added a reporting module to the administrator dashboard so that reports on student performance could
be generated and downloaded.
➢ Added data aggregation functions to compute metrics for individual student performance and class
averages.

❖ Testing & Quality Assurance:

➢ Carried out unit testing to verify that each component and feature of the system functions as intended.
➢ To guarantee smooth communication between the front end and back end, integration testing was
carried out.
➢ Conducted user acceptance testing (UAT) using test data to confirm the system's functionality and
performance for various user roles.

❖ Deployment & Documentation:

➢ Ensured that the application is available to all users by deploying it on a Firebase hosting environment.
➢ The complete system has been documented, including technical documentation for upcoming updates
and maintenance as well as installation guidelines and user manuals for educators, administrators, and
students.

❖ User Advice & Feedback:

➢ Offered administrators, teachers, and students training sessions to acquaint them with the system.
➢ Gathered user feedback to find any problems or potential areas for development, which will be fixed
in upcoming releases.

With this work completed, the process of efficiently managing student marks is streamlined and the Smart
Scan is guaranteed to be fully operational, secure, and ready for use by all stakeholders.

5.2 Conclusion

Modernizing and simplifying the academic process of managing student marks has been made possible by
Smart Scan. The system gives teachers, administrators, and students alike a reliable, safe, and easy-to-use
platform by utilizing web-based technologies like React and Firebase.

The system's thoughtfully crafted interfaces facilitate smooth communication between various user roles,
guaranteeing accurate and accountable mark updates, reviews, and approvals. Only authorized users are able
to access sensitive data thanks to the integration of Firebase Authentication, and Firebase Firestore offers a
dependable and scalable solution for managing and storing marks.

48
Role-specific dashboards improve the user experience by providing customized features, allowing
administrators to monitor the entire process, teachers to manage marks, and students to view their performance.
The approval workflow also protects the credibility of the academic records by promising that marks are
accurate and are only finalized after careful review.

The system's ability to produce thorough reports on student performance also adds a great deal of value by
facilitating data-driven decision-making while providing insights into academic trends and results.

All things considered, Smart Scan improves transparency, preciseness, and efficacy at all levels of academic
administration in addition to streamlining the marks management process. It is a complete solution that
satisfies the requirements of contemporary educational establishments, opening the door to a safer and more
efficient learning environment.

5.3 Environmental Benefits

By encouraging a more environmentally friendly method of managing academic records and minimizing the
need for conventional paper-based procedures, Smart Scan has a number of positive effects on the
environment. Some of the main benefits for the environment are listed below:
➢ Cutbacks on Paper Consumption: Traditionally, keeping track of student grades has required a large
amount of paper for distribution, recording, and storage. The system does away with the need for printed
grade sheets, mark registers, and other paper-based documents by digitizing this process. As a result,
there is a significant decrease in the amount of paper used, which contributes to the preservation of
forests and lowers the carbon footprint of paper production.

➢ Reduced Need for Physical Storage: Physical storage of paper records necessitates facilities with
lighting, climate control, and other resource-intensive features. Schools and other organizations can free
up physical space and lessen their environmental impact by switching to a digital system instead of
maintaining sizable archives of paper records.

➢ Energy Effectiveness: When it comes to managing and storing data, cloud-based services like Firebase
enable more efficient use of energy than on-premises servers. Because they use cutting-edge
technologies and renewable energy sources, large-scale data centers that host cloud services are
frequently more energy-efficient and use less energy overall.

➢ Decrease in Emissions from Transportation: There is no longer a need to physically move documents
between departments or schools when using a digital marks management system. The system can be
accessed remotely by administrators, teachers, and students, which minimizes the need for travel and the
corresponding greenhouse gas emissions.

➢ Promotion of Ecological Methods: Educational institutions set an example for sustainability by

implementing a digital approach to marks management, which in turn encourages other departments to

49
 explore environmentally friendly practices as well. This may result in the institution taking a more all-
encompassing approach to sustainability and adopting digital solutions on a larger scale.

5.4 Future Work Plans

Several improvements and new features can be added to the Automated Marks Capturing System as it
develops further to boost automation, efficiency, and user experience. In order to accomplish complete
automation of paper reading, integrating physical paper handling mechanisms is a crucial area of
development. A thorough work plan outlining these advancements is provided below:

1. Automated Paper Handling Integration:

• Goal:
 Using a roller-based paper feeding mechanism, completely automate the reading and
digitizing of physical documents, such as exam papers and answer sheets.
• Execution:
 Create and incorporate a roller system that can feed paper into an OCR (Optical Character
Recognition) scanner or another device automatically.
 Put the paper roller system and OCR software in sync to guarantee accurate and continuous
reading of paper-based documents.
• Anticipated Result:
 By incorporating automated paper rollers, the system will be able to process physical
documents with less assistance from humans. This will simplify the process of digitising
written marks and minimize errors in manual data entry.

2. Enhanced Data Extraction and OCR:

• Goal:
 Using cutting-edge OCR technology, increase the efficiency and accuracy of data extraction
from scanned documents.
• Implementation:
 Combine OCR technologies driven by AI that can identify handwritten text and transform
it into digital information.
 Teach the OCR system to correctly decipher various handwriting styles and common
formatting found on student response sheets.
 To verify accuracy, implement verification steps where the extracted data is cross-
referenced with current student records in Firebase Firestore.
• Anticipated Result:
 With improved OCR capabilities, digitizing paper-based marks will take less time and
effort, increasing the system's dependability and efficiency.

50
3. Extended Analytics and Reporting:
• Goal:
 Enlarging the administrator dashboard's reporting functionalities to offer more thorough
insights and analytics.
• Implementation:
 Create sophisticated data visualization tools that provide information on patterns in subject
difficulty, student performance, and teacher grading consistency.
 Utilize predictive analytics tools to project performance trends in the future by analyzing
historical data.
 Permit administrators to personalize reports so they can concentrate on data points or
student groups.
• Anticipated Result:
 Administrators will be able to make better decisions and more focused interventions
thanks to their increased understanding of academic performance.

4. Development of Mobile Applications:

• Goal:
 By creating a mobile application that replicates the features of the web-based system, we
hope to improve accessibility.
• Implementation:
 Create a user interface that is easy to use on smaller screens for administrators, teachers,
and students that is mobile-friendly.
 Set up push alerts to receive updates on report availability, approvals, and mark
submissions.
 Enhance the data synchronization process between Firebase and the mobile app to
guarantee offline access when needed and real-time updates.
• Anticipated Result:
 The utilization of a mobile application is expected to enhance user engagement by offering
convenient, mobile access to the system, especially for educators and students.

5. Security Enhancements:
• Goal:
 Constantly enhancing system security to keep private academic information safe from
unwanted access.
• Application:
 To provide an additional degree of protection, apply multi-factor authentication (MFA) to
every user role.
 To find and reduce possible risks, do frequent vulnerability assessments and security audits.
51
  Improve data transmission and storage encryption protocols to guarantee the safe handling
of all student grades and personal data.
• Anticipated Result:
 Stricter security protocols will guard against information leaks and guarantee the privacy
and accuracy of educational documents.

6. User Feedback and Iterative Improvements:

• Goal:
 To continuously improve the system in response to user input and new developments in
technology.
• Implementation:
 To get feedback on the performance and usability of the system, set up frequent feedback
loops with educators, administrators, and students.
Iteratively improve the system by implementing agile development techniques, with a focus
on bug fixes and improvements to the user experience.
• Anticipated Result:
 The system will remain relevant and efficient over the long run by staying in step with user
demands and technology developments thanks to ongoing improvement.

52
 APPENDIX A: REFERENCES

53
APPENDIX B: PLAGIARISM REPORT

54