MAHARASHTRA STATE BOARD OF TECHNICAL

EDUCATION, (MUMBAI)

A
Project Report
On

“Virtual Try-On”

Submitted by

• Purushottam Dilip Medhe

• Harshal Sanjay Kumawat
• Sarthak Yadav Sonar
• Dev Depak Pathak

Guided by: Mr. H. M. Gaikwad

Department of Artificial Intelligence and Machine Learning

K.K. WAGH POLYTECHNIC, NASHIK
(2023-2024)
 Karmaveer Kakasaheb Wagh Education Society, S
KARMAVEER KAKASAHEB WAGH POLYTECHNIC
Hirabai Haridas Vidyanagari, Amrutdham, Panchavati, Nashik-3

Certificate
Certified that the project report entitled “Virtual Try-On” has been
successfully completed by:
Name of student Class Enrolment No. Exam Seat No.
1. Purushottam Dilip Medhe TYAN 2100780311 399313
2. Harshal Sanjay Kumawat TYAN 2100780305 399305
3. Sarthak Yadav Sonar TYAN 2200780816 399343
4. Dev Depak Pathak TYAN 2100780317 399317
as partial fulfillment of Diploma course in Artificial Intelligence and Machine
Learning under the Maharashtra State Board of Technical Education, Mumbai
during the academic year 2023-2024.
The said work has been assessed by us and we are satisfied that the same
is up to the standard envisaged for the level of the course. And that the said
work may be presented to the external examiner.

H.M.Gaikwad H.M.Gaikwad
GUIDE HOD

External Examiner Prof. P. T. Kadave

PRINCIPAL

i
 Achievements / Participation Certificates

Runner Up at State Level Project Competition

ii
iii
Technophilia Project Competition

iv
v
 Acknowledgment

With deep sense of gratitude we would like to thanks all the people who have lit our path with
their kind guidance for our Project Selection, Design and Development. We are very grateful to these
intellectuals, experts, who did their best to help during our completion of project work.
It is our proud privilege to express deep sense of gratitude to, Prof. P. T. Kadave-Principal,
K. K. Wagh Polytechnic, Nashik for his comments and kind permission to complete this project. We
remain indebted to Prof. H. M. Gaikwad, Head of Artificial Intelligence & Machine Learning
Department for his timely suggestion and valuable guidance.
The special gratitude goes to our Internal Faculty Guide Mr. H. M. Gaikwad, staff members,
technical staff members, of Artificial Intelligence & Machine Learning Department for his/her
technical, timely, excellentand coercive guidance in completion of this project work. We thanks to
all the class colleagues for their appreciable, encouraging help for completion of our project.

We are also thankful to our parents who providing their wishful support for our project
completion successfully. Lastly we thanks to our all friends and the people who are directly or
indirectly related to our project work.

vi
 Vision & Mission

Institute Vision: Strive to empower students with Quality Technical Education.

Institute Mission: Committed to develop students as Competent and Socially Responsible Diploma
Engineers by inculcating learning to learn skills, values and ethics, entrepreneurial attitude, safe and
eco-friendly outlook and innovative thinking to fulfill aspirations of all the stakeholders and
contribute in the development of Organization, Society and Nation.

Department Vision: (Version – 1.2) To Develop Technocrats in the field of Computer Engineering
and Artificial Intelligence by imparting quality technical education.

Department Mission: (Version – 1.2)

M1: To provide quality technical education to students to help them to achieve good
academic growth.
M2: To impart technical education to meet the requirements of the industry and society.
M3: Develop technical & soft skill through co–curricular and extra-curricular activities for
improving personality.
Program Educational Objectives: (Version – 1.2)

PEO1: Provide IT and AI solutions to variety of industrial & social problems adapting
Profession ethics.
PEO2: Adapt state-of-the-art Computer Engineering and Artificial Intelligence broad-based
technologies to work in multi-disciplinary work environments.

PEO3: Solve broad-based problems individually and as a team member by effective

communication.

Program Specific Outcome: (Version – 1.2)

PSO1: Apply fundamental concepts of Computer Engineering, Artificial Intelligence and

Machine Learning to solve technical problems.
PSO2: Implement the domain knowledge to achieve successful career as an engineering
professional.

vii
 Course Outcomes: (COs)

CO607: Implement the planned activity individually and/or as team.

CO607.2: Select, collect and use required information/knowledge to solve the identified problem.
CO607.3: Take appropriate decisions based on collected and analyzed information.
CO607.4: Ensure quality in product.
CO607.5: Incorporate energy and environment conservation principles.
CO607.6: Consider the ethical issues related to the project (if there are any).
CO607.7: Assess the impact of the project on society (if there is any).
CO607.8: Communicate effectively and confidently as a member and leader of team.
CO607.9: Prepare project report after performing due plagiarism check using appropriate tools.

Program Outcomes: (POs)

PO1: Basic and Discipline specific knowledge: Apply knowledge of basic mathematics, science and
engineering fundamentals and engineering specialization to solve the engineering problems.
PO2: Problem analysis: Identify and analyze well-defined engineering problems using codified
standard methods.
PO3: Design/ development of solutions: Design solutions for well-defined technical problems and
assist with the design of systems components or processes to meet specified needs.
PO4: Engineering Tools, Experimentation and Testing: Apply modern engineering tools and
appropriate technique to conduct standard tests and measurements.
PO5: Engineering practices for society, sustainability and environment: Apply appropriate
technology in context of society, sustainability, environment and ethical practices.
PO6: Project Management: Use engineering management principles individually, as a team member
or a leader to manage projects and effectively communicate about well-defined engineering activities.
PO7: Life-long learning: Ability to analyze individual needs and engage in updating in the context of
technological changes.
Program Specific Outcomes (PSOs)
PSO1: Apply fundamental concepts of Computer Engineering, Artificial Intelligence and Machine
Learning to solve technical problems.
PSO2: Implement the domain knowledge to achieve successful career as an engineering professional.

COs, POs, and PSOs Mapping Table:

CO PO1 PO2 PO3 PO4 PO5 PO6 PO7 PSO1 PSO2

CO-1  
CO-2         
CO-3      
CO-4         
CO-5 
CO-6    
CO-7 
CO-8  
CO-9  
viii
 Abstract
Trying clothes in clothing stores is usually a time-consuming activity. Besides, it might not even be
possible to try-on clothes in such cases as online shopping. Our motivation here is to increase the time
efficiency and improve the accessibility of clothes try on by creating a virtual dressing room
environment. In this work, we introduce a virtual dressing room application using Microsoft Kinect
sensor. Our proposed approach is mainly based on extraction of the user from the video stream,
alignment of models and skin color detection. We use the modules for locations of the joints for
positioning, scaling and rotation in order to align the 2D cloth models with the user. Then, we apply skin
color detection on video to handle the unwanted collusions of the user and the model. Finally, the model
is superimposed on the user in real time. The problem is simply the alignment of the user and the cloth
models with accurate position, scale, rotation and ordering. First, detection of the user and the body parts
is one of the main steps of the problem. In literature, several approaches are proposed for body part
detection, skeletal tracking and posture estimation, and superimposing it onto a virtual environment in
the user interface. The project is implemented in C# programming environment for real time, Kinect
hacking application. Kinect driver’s middleware are used for various fundamental functions and for the
tracking process in combination with Kinect sensor.

Keywords: Virtual dressing room, Microsoft Kinect, skeletal tracking.

ix
 Table of Contents

Sr. No. Name of Topic Page No.

Certificates I
Achievements/Participation certificate II
Acknowledgment VI
Vision and Mission VII
CO & PO Mapping VIII
Abstract IX
Table of contents X
1. Introduction
1.1. Literature Survey 1
1.2. Existing System Overview 2
1.3. Proposed Concept and Working 3
1.4. Proposed System Architecture 3
2. Project Requirements and feasibility analysis

2.1. Overview of Proposed Project 5

2.2. Scope and Area of Implementation 5
2.3. Software Requirement Specification (SRS) 6
2.3.1 Functional Requirements 6
2.3.2 Non-Functional Requirement 7
2.4 Software Requirements 7
2.5 Hardware Requirements 8
2.6 Preliminary Survey of Virtual Try-On Systems 9
2.6.1Operational Feasibility Assessment 10
2.6.2 Technical Feasibility Evaluation 10
2.6.3 Economic Feasibility Analysis 11
3. Project Design and Implementation 13
3.1 System Architecture Diagram 13
3.2 Data Flow Diagram (DFD) 14
3.2.1 DFD Level 0 14
3.2.2 DFD Level 1 14
3.3 Unified Modeling Language (UML) Diagram 14
3.3.1 Use Case Diagram 15
3.3.2 Sequence Diagram 17
3.3.3 Activity Diagram 18

x
 3.4 Class Diagram 19
3.5 Module Analysis and Description 20
4. Results and Outputs/ Screenshots 26
4.1 Virtual Try-On Interface 26
4.2 Menu 27
4.3 Male Clothes Section 29
4.4 Female clothes section 30
5. Testing of software application 32
5.1 Unit Testing of Individual Modules 32
5.2 Integration Testing 32
5.3 User Acceptance Testing (UAT) with Kinect Sensor 33
6. Costing Of Project 35
7. Application of virtual try-on system 36
7.1 Retail Industry 36
7.2 Fashion Designers and Brands 36
7.3 Online Marketplaces 36
7.4 Personal Styling Services 37
8. Future Scope 38
8.1 Deeper Integration with AI and Machine Learning 38
8.2 Expanding the virtual Wadrobe 38
8.3 Integration with augumented Reality 38
8.4 Advanced Data Analytics 38
9. Conclusion 39
10. References 40

xi
 Index of Figures

Fig. No. Name of figure Page no.

3.1 System Architecture Diagram 13

3.2.1 DFD Level 0 diagram 14

3.2.2 DFD Level 1 Diagram 14
3.3.1 Use Case Diagram 15
3.3.2 Sequence Diagram 17
3.3.3 Activity Diagram 18
3.4 Class Diagram 19
4.1 Virtual Try-On Interface 26

4.2 Virtual Try-On Menu 27

4.3 Virtual Try-On Male Clothes Section 29
4.4 Virtual Try-On Female Clothes Section 30

xii
………………………………………………………………………. Virtual Try-On…………

Chapter 1
INTRODUCTION

1. Introduction
This project aims to revolutionize the shopping experience by allowing users to virtually try on
various clothes and accessories. Through the use of advanced image processing and augmented
reality techniques, users can visualize how different garments look on them without physically
trying them on. Our project holds the ability to transform the way people shop for clothes and
accessories. Not only does this enhance the online shopping experience, but it also saves time
and reduces the need for physical try-ons. Let's dive in and explore the exciting world of virtual
fashion.

1.1. Literature Survey

1. Lenskart:
• Description: Lenskart, an eyewear retailer, offers a virtual try-on feature through its website and
mobile app. Users can upload a photo or use their device's camera to virtually try on different
styles of eyeglasses and sunglasses. The system uses augmented reality technology to
superimpose virtual glasses onto the user's face, allowing them to see how each frame looks in
real-time. Lens kart’s virtual try-on feature also includes tools for adjusting frame fit, color, and
lens options.
• Literature Review: Research on virtual try-on systems for eyewear, such as Lenskart's, has
examined the impact of virtual try-on technology on consumer purchase behaviour and
satisfaction. Studies have investigated factors influencing the perceived accuracy and realism of
virtual try-on experiences, including frame alignment, lens distortion correction, and user
interface interactivity. Additionally, research has explored the role of virtual try-on systems in
reducing the uncertainty associated with online eyewear purchases and increasing user
confidence in selecting frames.

2. L'Oréal Makeup Genius:

• Description: L'Oréal, a leading cosmetics company, offers a virtual try-on app called Makeup
Genius. Users can use their smartphone or tablet camera to virtually try on a wide range of
makeup products, including lipstick, eyeliner, and blush. The app uses augmented reality
technology to accurately map facial features and apply virtual makeup in real-time. Users can
experiment with different makeup looks, customize colors and styles, and share their virtual
makeovers on social media.
• Literature Review: Research on virtual try-on systems for cosmetics, like L'Oréal Makeup
Genius, has explored the impact of virtual makeup try-on experiences on consumer engagement
and brand loyalty. Studies have investigated user preferences for virtual makeup customization
options, such as color selection, texture application, and virtual makeover sharing features.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 1 of 37
………………………………………………………………………. Virtual Try-On…………

3. Warby Parker:
• Description: Warby Parker, an eyewear retailer, offers a virtual try-on feature on its website and
mobile app. Users can use their device's camera to virtually try on different styles of glasses and
sunglasses. The system uses facial recognition technology to accurately position the virtual
glasses on the user's face, allowing them to see how each style looks from various angles.
• Literature Review: Studies examining virtual try-on systems for eyewear, such as Warby
Parker's, have explored the impact of virtual try-on experiences on consumer purchase behaviour
and decision-making. Research has investigated factors influencing the perceived realism of
virtual try-on experiences, including accurate rendering of frames, facial tracking precision, and
user interface design. Additionally, studies have evaluated the effectiveness of virtual try-on
systems in reducing the perceived risk of online eyewear purchases and increasing user
confidence in selecting frames.

4. Sephora Virtual Artist:

• Description: Sephora, a cosmetics retailer, offers a virtual try-on feature called Virtual Artist on
its website and mobile app. Users can virtually try on different makeup products, including
lipstick, eyeshadow, and foundation, using augmented reality technology. The system analyzes
facial features and skin tones to provide accurate color matching and realistic makeup
simulations.
• Literature Review: Research on virtual try-on systems for cosmetics, like Sephora's Virtual
Artist, has examined the role of augmented reality in enhancing the online shopping experience
for beauty products. Studies have investigated consumer attitudes and preferences towards
virtual makeup try-on experiences, as well as the impact of virtual try-on technology on purchase
behaviour and brand loyalty. Additionally, research has explored the technical challenges of color
accuracy, texture mapping, and facial feature detection in virtual makeup try-on systems.

1.2. Existing System Overview

The existing system utilizes physical fitting rooms. Customers locate desired items on the
shopping floor and take them to fitting rooms. This approach has limitations such as:

• Limited availability: Fitting rooms can become crowded during peak hours, leading to waiting
times and customer frustration.
• Inefficiency: Customers may need assistance from staff to retrieve different sizes or styles,
impacting staff workload.
• Lack of customization: The current system doesn't allow customers to experiment with various
combinations of clothing or visualize different looks."
By outlining the limitations of the existing system, you can effectively highlight the potential
benefits that VTO technology can bring to your store.

Additional Considerations
• If your store uses any basic tools like size charts or recommendation mirrors, you can briefly
mention them here.
==================================================================
K. K. Wagh Polytechnic, Nashik Page 2 of 37
………………………………………………………………………. Virtual Try-On…………

• Focus on how the existing system creates friction or limitations in the customer experience.
• Ensure a smooth transition by acknowledging the existing system and then presenting VTO as a
solution for improvement.

1.3. Proposed Concept and Working

Concept
Virtual Try-On System using computer vision and augmented reality technologies. The primary
goal is to create a platform that allows users to virtually try on clothing and accessories in an
immersive and realistic manner without physically wearing the items.
Working
The Virtual Try-on application begins with a person standing in front of a fixed distance Kinect.
The Kinect analyzes the person's body, identifying key points and calculating necessary
distances.
User can then resize and try on the garments, with the application suggesting appropriate sizes
and offering stylistic tips.

1.4. Proposed System Architecture

1. Microsoft Kinect Sensor Integration:

• Utilize Kinect's depth-sensing technology to capture precise 3D spatial information of the user's
body.
• Leverage Kinect's skeletal tracking capabilities to accurately detect and track the user's body
movements and poses.
• Integrate Kinect sensor with the virtual try-on system to provide real-time feedback on user
interactions.

2. Real-time Body Tracking and Motion Capture:

• Implement algorithms to process depth data from the Kinect sensor and extract skeletal
information, including joint positions and orientations.
• Continuously track the user's body movements and gestures with high accuracy and low latency.
• Translate the captured skeletal data into virtual avatar animations for seamless visualization of
clothing items on the user's body.

3. Garment Simulation and Visualization:

• Develop garment simulation algorithms to realistically drape virtual clothing items onto the
user's virtual avatar.
• Utilize physics-based cloth simulation techniques to accurately model fabric behavior and
simulate realistic garment movements.
• Incorporate rendering techniques to achieve high-fidelity visualizations of virtual garments,
including textures, colors, and material properties.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 3 of 37
………………………………………………………………………. Virtual Try-On…………

4. User Interaction and Interface:

• Design an intuitive user interface for interacting with the virtual try-on system, integrating
Kinect's gesture recognition capabilities for natural and intuitive interactions.
• Enable users to select and customize virtual clothing items through gesture-based or voice-based
commands, leveraging Kinect's input modalities.
• Provide visual feedback to users during the try-on process, allowing them to adjust garment
placement, size, and orientation in real-time.

5. Machine Learning for Personalization:

• Integrate machine learning algorithms to analyse user preferences, body measurements, and style
preferences.
• Develop personalized recommendation systems to suggest clothing items based on user profiles
and previous interactions.
• Utilize machine learning models to adapt garment fitting and styling based on individual user
characteristics and feedback.

6. Integration with E-commerce Platforms:

• Enable seamless integration with e-commerce platforms to facilitate online shopping and
purchase transactions.
• Synchronize virtual try-on experiences with product catalogues, inventory databases, and pricing
information from e-commerce databases.
• Implement features for adding selected items to the shopping cart, viewing product details, and
completing the purchase process within the virtual try-on environment.

7. Scalability and Performance Optimization:

• Optimize the virtual try-on system for performance and scalability to support large user bases
and concurrent usage.
• Implement parallel processing techniques to efficiently handle real-time data streams from
multiple Kinect sensors.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 4 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 2
PROJECT REQUIREMENTS AND FEASIBLITY ANALSYSIS

2.1 Overview of Proposed Project

This project proposes a virtual try-on system designed to transform the online shopping experience.
The system is built with a C#/ASP.NET (.NET) frontend and a Python backend. The frontend of the
platform is built using the C#/ASP.NET (.NET) framework, while the backend is developed using
Python.

Leveraging the power of computer vision and augmented reality (AR), we aim to create a platform
that allows users to virtually try on clothing and accessories in an immersive and realistic way.
Imagine browsing online stores and being able to see how clothes would look on you without ever
having to step into a fitting room. This innovative system utilizes cutting-edge technologies:

• Computer Vision: Sophisticated algorithms will analyse a user's body pose and dimensions through
their webcam or smartphone camera.
• Augmented Reality: The chosen clothing or accessory will be virtually overlaid onto the user's
image, providing a realistic representation of how the item would fit and appear.

By providing this virtual try-on experience, we hope to achieve several key benefits:
• Enhanced Customer Confidence: Users can shop online with greater confidence, knowing exactly
how clothes will look on them before they buy.
• Reduced Return Rates: The ability to virtually try-on items can significantly reduce the number of
returns for e-commerce businesses.
• Improved User Experience: This innovative approach will make online shopping more engaging,
interactive, and enjoyable for users.

Overall, our proposed project has the potential to empowers informed clothing purchases, boosting
confidence and potentially reducing retailer returns. This innovative approach, with effective
implementation and marketing, has the potential to reshape online fashion shopping.

2.2 Scope and Area of Implementation

The virtual try-on system we propose has the potential to be implemented across various segments of
the retail industry:
• E-commerce fashion retailers: This technology can be seamlessly integrated into online stores,
allowing customers to virtually try on clothing and accessories before purchasing, leading to
increased sales and reduced return rates.
• Brick-and-mortar retail: In-store kiosks equipped with virtual try-on features can enhance the
customer experience, offering a fun and interactive way to try on clothes without needing
physical fitting rooms.
• Luxury brands: Virtual try-on can provide a more personalized and engaging experience for
luxury shoppers, allowing them to virtually try on high-end items before visiting a physical store.
==================================================================
K. K. Wagh Polytechnic, Nashik Page 5 of 37
………………………………………………………………………. Virtual Try-On…………

1. In-Store Virtual Try-On Stations:

o Scope: Set up dedicated stations within the store where customers can virtually try on
clothes.
o Implementation Areas:
▪ Hardware: Install cameras, screens, and sensors for real-time interaction.
▪ Software: Develop or integrate virtual try-on software.
▪ User Interface: Design an intuitive interface for customers.
▪ Space Allocation: Allocate space within the store for these stations.

2. Privacy and Hygiene Considerations:

o Scope: Address privacy concerns related to capturing customer images.
o Implementation Areas:
▪ Privacy Policies: Clearly communicate data handling practices.
▪ Hygiene: Regularly sanitize cameras and screens.

3. Marketing and Promotion:

o Scope: Promote the availability of virtual try-on in-store.
o Implementation Areas:
▪ Signage: Display clear signage indicating the presence of virtual try-on stations.
▪ Social Media: Share posts and stories about the technology.

2.3 Software Requirement Specification (SRS)

2.3.1 Functional Requirements

• User Interface: The user interface will allow users to browse through a catalog of clothing and
accessory items, select items for virtual try-on, and see their live image from the webcam or
smartphone camera.
• Computer Vision: The system will utilize computer vision algorithms to accurately detect and
analyze the user's body pose and dimensions.
• Augmented Reality Integration: The chosen clothing or accessory item will be virtually
overlaid onto the user's image in real-time using AR technology. This overlay should provide a
realistic representation of how the item would fit and appear on the user's body.
• Pose Estimation: The system should be able to adapt to different user poses (standing, sitting,
etc.) to ensure an accurate virtual try-on experience.
• Size and Style Variations: Users should be able to select different sizes and styles for the virtual
try-on items.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 6 of 37
………………………………………………………………………. Virtual Try-On…………

2.3.2 Non-functional Requirements

• Performance: The virtual try-on experience should be smooth and responsive, with minimal
latency between user actions and visual updates.
• Accuracy: The system should accurately detect and analyze user body dimensions to provide a
realistic virtual try-on experience.
• Security: User privacy should be a priority. The system should handle user data securely and
comply with relevant data protection regulations.
• Usability: The user interface should be intuitive and easy to navigate, allowing users to easily
explore the virtual try-on functionalities.
• Scalability: The system should be scalable to accommodate a growing user base and future
feature additions.

2.4 Software Requirements

• Visual Studio Code: A source-code editor made by Microsoft with the Electron Framework, for
Windows, Linux and macOS.

• Blender: Free and open-source 3D creation suite for modeling, sculpting, animating, simulating,
rendering, and compositing 3D assets (potentially useful for creating virtual clothing models for
try-on).

• Autodesk Maya: Industry-standard professional 3D animation, modeling, simulation, and

rendering software (similar to Blender, but with a wider range of advanced features).

• OpenCV: Open Source Computer Vision Library offering a comprehensive set of algorithms for
real-time computer vision tasks (potentially used for user pose estimation in your virtual try-on
system).

• Firebase: Google's mobile development platform providing a suite of backend services like
databases, authentication, storage, analytics, and machine learning (could be used to manage user
data, store virtual try-on preferences).

2.5 Hardware Requirements

Camera sensors: Microsoft Kinect (Xbox 360 1920*1080,30 fps)

Display: LCD, LED panel
Storage: 512 GB SDD (may vary according to CatLog size)
Operating System: Windows 11
RAM: 8GB or above
CPU: i5 13th gen

==================================================================
K. K. Wagh Polytechnic, Nashik Page 7 of 37
………………………………………………………………………. Virtual Try-On…………

Microsoft Kinect Sensor

The Microsoft Kinect for Xbox 360, a revolutionary motion sensing input device, boasts a depth sensor
with a resolution of 640x480 at 30 frames per second (fps) and a colour sensor of the same specifications,
providing a comprehensive understanding of the user's environment. The device excels in skeleton
tracking, recognizing multiple users simultaneously and associating them with individual profiles. It
further extends its capabilities through facial recognition. Compatible with both Xbox 360 and Windows
PC, Kinect's software development kit (SDK) empowers developers to harness its potential in various
applications, ranging from human-computer interaction to virtual reality and robotics.

Specifications:
1. Depth Sensor: 640x480 resolution at 30 frames per second (fps)
2. Colour Sensor: 640x480 resolution at 30 fps
3. Field of View: Horizontal field of view is 57 degrees, vertical field of view is 43 degrees
4. Audio: Four microphone array
5. Connectivity: USB 2.0
6. Power: AC adapter with USB connector
7. Dimensions: 250 x 75 x 88 mm (9.8 x 3.0 x 3.5 in)
8. Weight: 275 grams (9.7 oz)
9. Compatibility: Xbox 360, Windows PC (with appropriate drivers and software)
10. Operating Range: Approximately 1.2m to 3.5m (4 ft to 12 ft)

Features:
1. Motion Sensing: Kinect can detect the movement of users in its field of view, allowing for
gesture-based control.
2. Depth Sensing: Utilizes infrared technology to create a depth map of the scene, enabling
accurate tracking of objects and users in three dimensions.
3. Voice Recognition: Incorporates a microphone array for voice commands, allowing users to
interact with the system using speech.
4. Skeleton Tracking: Can identify and track multiple users simultaneously, recognizing their
bodies and movements to enable interactive experiences.
5. Facial Recognition: Capable of recognizing faces and associating them with user profiles.
6. Software Development Kit (SDK): Provides developers with tools and libraries to create
applications and experiences utilizing Kinect's capabilities.
==================================================================
K. K. Wagh Polytechnic, Nashik Page 8 of 37
………………………………………………………………………. Virtual Try-On…………

Project Details:
1. Objective: Clearly state the purpose of using Kinect in your project. Whether it's for human-
computer interaction, gesture recognition, virtual reality, robotics, or any other application.
2. Implementation: Describe how you integrated Kinect into your project setup. Detail any
hardware configurations and software libraries used.
3. Functionality: Explain the functionalities enabled by Kinect in your project. For example, if
you're using it for gesture recognition, describe the gestures recognized and corresponding
actions.
4. Challenges: Discuss any challenges faced during the implementation process, such as calibration
issues, noise interference, or limitations of the Kinect sensor.
5. Results and Evaluation: Present the results of your project and evaluate its performance.
Include any quantitative metrics or qualitative assessments.
6. Future Work: Propose potential enhancements or future directions for your project. This could
involve improving accuracy, expanding functionality, or exploring new applications.

2.6 Preliminary Survey of Virtual Try-On Systems

Real time video-Based Virtual Try-On:

o These systems use computer technology to assist users in trying on garments virtually.
o Users can see how clothes look on themselves, how different pieces match together, and
how the size fits.
o Deep learning methods play a crucial role in developing these systems.
o Challenges include accurate body representation, realistic cloth simulation, and efficient
rendering.
2.6.1 Operational Feasibility Assessment

1. What Is Virtual Try-On?

o Virtual try-on refers to digitally trying on garments or accessories in a virtual
environment.
o Users can visualize how clothing items look on themselves or an avatar without
physically trying them on.
o This technology aims to address challenges associated with conventional e-commerce,
such as fit issues and return costs.

2. How Does Virtual Try-On Work?

o Virtual garment try-ons come in two formats:
▪ At-Home Try-On: Users try on items via the internet at home using 3D virtual
fitting technology and a camera-equipped device.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 9 of 37
………………………………………………………………………. Virtual Try-On…………

▪ In-Store Try-On: Customers visit physical stores where they can virtually try on
clothes.
o The process involves computer vision, artificial intelligence, recommendation
algorithms, and augmented reality (AR) or virtual reality (VR).
o 3D body scanning combined with AR technology can revolutionize the fashion store
experience.

3. Advantages of Virtual Try-On:

o Improving Personalization: Tailored recommendations enhance the shopping
experience.
o No-Contact Try-On: Especially relevant during pandemic times, virtual try-on avoids
physical contact with clothes.

2.6.2 Technical Feasibility Evaluation

1. Hardware and Infrastructure:

o Computing Power: Ensure that the store’s hardware (computers, GPUs, etc.) can handle
real-time rendering and interaction with virtual clothing.
o Cameras and Sensors: Install high-resolution cameras or depth sensors to capture users’
movements accurately.
o Network Connectivity: Reliable internet connectivity is essential for seamless
interactions.
o
2. Software and Algorithms:
o 3D Modeling and Rendering: Develop or choose software capable of creating realistic
3D models of clothing items.
o Cloth Simulation: Implement physics-based cloth simulation algorithms to ensure
accurate draping and movement.
o Real-Time Interaction: Design responsive interfaces for users to interact with virtual
garments.
3. User Experience:
o Latency: Minimize latency between user actions and system responses to maintain a
natural experience.
o Realism: Strive for realistic clothing appearance, including texture, lighting, and
shadows.
o User Interface: Create an intuitive interface for users to select, try on, and adjust virtual
clothes.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 10 of 37
………………………………………………………………………. Virtual Try-On…………

4. Integration with Existing Systems:

o Point of Sale (POS): Integrate virtual try-on with existing POS systems for seamless
transactions.
o Inventory Management: Ensure that virtual clothes correspond to actual inventory.
o Security and Privacy: Address data privacy concerns related to user images and
measurements.

5. Scalability and Maintenance:

o Scalability: Consider how the system will handle multiple users simultaneously.
o Maintenance: Regularly update software, monitor hardware health, and address any
technical issues.
6. User Training and Support:
o Staff Training: Train store staff to assist users with the virtual try-on process.
o User Guidance: Provide clear instructions for users on how to use the system.

2.6.3 Economic Feasibility Analysis

1. Costs and Investments:
o Initial Setup: Evaluate the expenses related to hardware (computers, cameras, sensors),
software development, and integration with existing systems.
o Maintenance and Upgrades: Consider ongoing costs for system maintenance, updates,
and technical support.
2. Return on Investment (ROI):
o Calculate the potential ROI by comparing the expected benefits (reduced returns,
improved customer experience) with the implementation costs.
o Assess how quickly the investment pays off based on increased sales and operational
efficiency.
3. Operational Efficiency:
o Virtual try-on systems can streamline the shopping process, reduce the need for physical
fitting rooms, and enhance customer satisfaction.
o Evaluate the impact on staff workload and overall store operations.
4. Customer Experience and Sales Impact:
o Virtual try-on can attract more customers, increase engagement, and boost sales.
o Consider how it aligns with your store’s brand and customer expectations.
5. Competitive Advantage:
o Assess whether offering virtual try-on sets your store apart from competitors.
o Consider how it enhances your brand image and attracts tech-savvy customers.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 11 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 3
PROJECT DESIGN & IMPLEMENTATIONS

3.1 System Architecture Diagram

Fig 3.1: System Architecture Diagram

• Kinect Sensor: This depth camera captures a three-dimensional image of the user. It provides
more information than a regular webcam, allowing for a more accurate estimation of the user's
body pose.
• Positioning and Measurements: This block likely refers to the process of using the Kinect
sensor data to determine the user's position and body measurements in 3D space.
• Look for dress in database: This block searches a database of clothing items for a dress that
matches the user's body dimensions (potentially based on the measurements obtained from the
Kinect sensor).
• 3D dress Trial: This block represents the virtual try-on experience itself. The system retrieves a
3D model of the dress from the database and virtually drapes it onto the user's 3D body model
(created from the Kinect data).
• LCD Screen: This displays the final virtual try-on image or video, showing the user how the
dress would look on them in real-time.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 12 of 37
………………………………………………………………………. Virtual Try-On…………

3.2 Data Flow Diagram (DFD)

3.2.1 DFD Level 0

Fig. 3.2.1: DFD Level 0 diagram

User: Represents the input provider.

Virtual Dressed Up System: Processes the user's input and creates the dressed-up result.
Result: Displays or presents the dressed-up user to the user.

3.2.2 DFD Level 1

Fig 3.2.2: DFD Level 1 Diagram

1. User: The user interacts with the system and provides input data, such as images or videos.
2. Virtual Dressed-Up System: This process takes user input and creates a virtual dressed-up view
based on the provided clothing items.
3. Video: An external source of video input, which is part of the user's input.
4. Analysis: It takes the video input and performs several tasks, as described in explanation.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 13 of 37
………………………………………………………………………. Virtual Try-On…………

3.3 Unified Modeling Language (UML) Diagram

3.3.1 Use Case Diagram

Fig. 3.3.1: Use Case Diagram

Use case diagram is a graph of actors, a set of use cases enclosed by a system boundary, communication
associations between the actor and the use case. The use case diagram describes how a system
interacts with outside actors; each use case represents a piece of functionality that a system
provides to its users. A use case is known as an ellipse containing the name of the use case and an actor
is shown as a stick figure with the name of the actor below the figure.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 14 of 37
………………………………………………………………………. Virtual Try-On…………

The diagram shows the following steps in the virtual try-on process:

1. Input Video: The system takes a video input from the user. This video is presumably of the user
themselves.
2. Convert into Frames: The video is then divided into individual frames.
3. Apply Euclidean Distance Formula: A mathematical formula is applied to each frame to likely
calculate the distance between different points on the user's body. This helps create a body map.
4. Shrink Image: The image frames are likely resized to a smaller dimension to reduce processing
power needed.
5. Build Model: A model is built, possibly a 3D model, based on the processed image frames.
6. Prediction: The system predicts how clothing would look on the user's body based on the 3D
model.
7. Result: The final result is displayed to the user, which is likely an image or video of the user
with virtual clothing on.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 15 of 37
………………………………………………………………………. Virtual Try-On…………

3.3.2 Sequence Diagram

Fig. 3.3.2: Sequence Diagram

The above sequence diagram shows the process of the virtual try-on. Here's a breakdown of the
steps involved:

1. Input Video: The first step involves the user inputting a video.
2. Load Data: The system loads the video data for processing.
3. Convert into Frames: The video is then divided into individual frames.
4. Image Classification: An image classification process is performed, likely to identify different
parts of the user's body in each frame.
5. Build Model: A 3D model of the user's body is built based on processed image frames.
6. Check for Accuracy: The accuracy of the 3D model is evaluated.
7. Classification: If the accuracy is acceptable, the classification process is likely used to overlay
virtual clothing onto the 3D model.
8. Output: The final result, which is an image or video of the user with virtual clothing on, is
displayed.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 16 of 37
………………………………………………………………………. Virtual Try-On…………

3.3.3 Activity Diagram

Fig. 3.3.3: Activity Diagram

The above activity diagram illustrates the steps involved in the image pre-processing stage
of the virtual try-on. Here's a breakdown of the process:

1. Datasets: The process begins with datasets, which are collections of images used to train the
system.

2. Pre-processing: The images undergo pre-processing, which involves several steps including
resizing, image analysis, and feature engineering.

3. Image Pre-processing Steps:

• Resize: The images are resized to a standard dimension.

• Image Analysis: An analysis is performed on the images, possibly to identify key points
like body landmarks.
==================================================================
K. K. Wagh Polytechnic, Nashik Page 17 of 37
 ………………………………………………………………………. Virtual Try-On…………

• Feature Engineering: New features are extracted from the images that are relevant to
the task of virtual try-on, such as body measurements or posture.

4. Validation: The pre-processed images are validated to ensure they meet the required quality
standards.

5. Output: The validated pre-processed images are then used for the next stage of the virtual try-
on process.

3.4 Class Diagram

Fig. 3.4: Class Diagram

The above class diagram illustrates the classes and their relationships in a virtual try-on system.
Here's a breakdown of the classes and their functionalities:

• User: This class represents the system's user and likely has attributes such as username, name, and
preferences.

• Garment: This class represents the clothing items that can be virtually tried on. It likely has attributes
such as garment ID, name, size, color, type (e.g., dress, shirt), and potentially a list of images.

• BodyPose: This class represents the user's body posture in an image or video frame. It might have
attributes like a list of body landmarks (e.g., wrist, ankle) and their corresponding positions.

• GarmentFitting: This class likely handles the logic of virtually trying on garments on the user's body.
It has methods like drapeGarment that takes a BodyPose object and a Garment object as inputs and
returns a VirtualTryOnResult object.
==================================================================
K. K. Wagh Polytechnic, Nashik Page 18 of 37
 ………………………………………………………………………. Virtual Try-On…………

• VirtualTryOnResult: This class likely represents the result of the virtual try-on process. It might have
methods to display the result (image or video) and potentially information about the fit of the garment.

• GarmentDatabase: This class stores information about the garments in the system. It likely has
methods for searching garments based on criteria like type, size, or color.

• BodyLandmark: This class likely represents a specific landmark on the user's body, such as a wrist or
ankle. It might have attributes like its name and position (x, y coordinates).

• UserInterface: This class likely represents the user interface of the virtual try-on system. It might
have methods for displaying garment lists, receiving user input, and showing the virtual try-on results.

3.5 Module Analysis and Description

1. Module Name: User Interface Module

• Purpose: Provides a user-friendly interface for interacting with the virtual try-on system.

• Inputs and Outputs:

o Inputs: User actions like selecting garments, capturing images/videos, providing size
information (optional).
o Outputs: Displays garment selection options, virtual try-on results, and potentially size
recommendation suggestions.

• Used by Module: C# and .NET

• Algorithm:
1. Presents a user interface for browsing garments.
2. Allows users to select garments and potentially provide additional details.
3. Provides options for capturing user images/videos (optional).
4. Displays the virtual try-on results after processing.
5. Offers functionalities like zooming, rotating the virtual garment (optional).

2. Module Name: Image/Video Input Module

• Purpose: Handles user-provided images or video for the virtual try-on process.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 19 of 37
………………………………………………………………………. Virtual Try-On…………

• Inputs and Outputs:

o Inputs: User interaction to capture an image or video using the device camera.
o Outputs: Captured image or video data passed to the processing pipeline.
• Used by Module: Device camera APIs or multimedia libraries (Microsoft.Kinect)

• Algorithm:
1. Provides an interface for users to capture an image or video of themselves.
2. Uses device camera APIs to capture the image/video data.
3. Saves or transmits the captured image/video data for further processing.

3. Module Name: Virtual Try-On Engine Module

• Purpose: Performs the core virtual try-on functionality, simulating how garments would look on
the user.

• Inputs and Outputs:

o Inputs: Selected garment data (image, size), user image/video (optional), user pose
information (optional).
o Outputs: Processed image or video showing the virtual garment on the user.

• Used by Module: Image processing libraries (e.g., OpenCV, Pillow) - 3D modeling frameworks
(e.g., Blender, Three.js) - Garment fitting algorithms.

• Algorithm:
1. Receives user input (garment selection, size) and potentially user image/video or pose data.
2. Processes user data (e.g., pose estimation from image/video).
3. Uses garment data and user information to simulate garment drape on the user's body (using
image manipulation or 3D modeling techniques).
4. Generates the final image or video representation of the virtual try-on experience.

4. Module Name: Image Analysis Module

==================================================================
K. K. Wagh Polytechnic, Nashik Page 20 of 37
………………………………………………………………………. Virtual Try-On…………

• Purpose: Analyzes user-provided images or videos to extract relevant information for the virtual
try-on process.

• Inputs and Outputs:

o Inputs: User-provided image or video.
o Outputs: Extracted information like user pose (joint positions), background
segmentation (separating user from background).

• Used by Module: Computer vision libraries (e.g., OpenCV) - Machine learning models for pose
estimation and image segmentation

• Algorithm:
1. Takes user image/video as input.
2. Applies image processing techniques to extract user pose information (e.g., joint locations).
3. Optionally performs background segmentation to isolate the user from the background
image.
4. Outputs the extracted information for further processing by the virtual try-on engine.

5. Module Name: Database Module

• Purpose: Stores information about available garments for the virtual try-on system.

• Inputs and Outputs:

o Inputs: New garment data (images, size charts, descriptions) might be added to the
database (depending on system design).
o Outputs: Provides garment data (images, size charts) to other modules upon request.

• Used by Module: Database management system (e.g., Firebase)

• Algorithm:
1. Provides functionalities to add new garment information to the database (if applicable).
2. Responds to requests from other modules, retrieving relevant garment data (images, size
charts) based on garment selection.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 21 of 37
………………………………………………………………………. Virtual Try-On…………

6. Module Name: Result Display Module

• Purpose: Presents the virtual try-on results to the user.

• Inputs and Outputs:

o Inputs: Processed image or video from the virtual try-on engine.
o Outputs: Displays the generated image or video showing the virtual garment on the user.

• Used by Module: Device camera APIs or multimedia libraries (Microsoft.Kinect)

• Algorithm:
1. Receives the processed image or video data from the virtual try-on engine.
2. Uses UI elements or multimedia players to display the virtual try-on result to the user.
3. Optionally offers functionalities like zooming, rotating the virtual garment on the
displayed image/video.
7. Module Name: User Interaction Module

• Purpose: Handles user interaction with the virtual try-on system beyond basic UI elements.

• Inputs and Outputs:

o Inputs: User gestures, clicks, selections within the virtual try-on interface.
o Outputs: Triggers actions within the system based on user interactions.

• Used by Module: Device camera APIs or multimedia libraries (Microsoft.Kinect)

• Algorithm:
1. Listens for user interactions within the virtual try-on interface (e.g., selecting different
garments, adjusting size options).
2. Translates user interactions into specific actions within the system (e.g., fetching new
garment data, re-running the virtual try-on with different parameters).
3. Provides feedback to the user based on their interactions (e.g., updating the display with new
results).

8. Module Name: Analytics Module

• Purpose: Tracks user interactions and virtual try-on behavior for analysis and improvement.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 22 of 37
………………………………………………………………………. Virtual Try-On…………

• Inputs and Outputs:

o Inputs: User interactions, virtual try-on session data (garment selections, results viewed).
o Outputs: Aggregated data and insights for analysis.

• Used by Module: Database management system (e.g., Firebase) and Microsoft.Kinect

• Algorithm:
1. Collects data on user interactions and virtual try-on sessions (e.g., selected garments, viewed
results).
2. Stores or transmits the collected data for further analysis.
3. Optionally generates reports or insights on user behavior and virtual try-on trends.

9. Module Name: Augmented Reality Module

• Purpose: Overlays the virtual garment onto a live video stream of the user for a more immersive
experience.
• Inputs and Outputs:
o Inputs: Live video stream from the device camera, user pose information.
o Outputs: Augmented reality view with the virtual garment overlaid on the user's live
video feed.
• Used by Module: AR frameworks (e.g., ARKit, ARCore) - Computer vision libraries for real-
time processing
• Algorithm:
1. Captures a live video stream from the device camera.
2. Estimates user pose information from the video stream.
3. Overlays the virtual garment onto the live video feed in real-time, considering user
pose and movement.
4. Displays the augmented reality view with the virtual garment on the user.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 23 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 4
RESULT AND OUTPUTS/ SCREEN SHOTS

4.1. Platform/ Interface

Fig.4.1. Virtual Try-On Interface

This image showcases the initial interface of the VTO system. The screen displays a prominent 'MENU'
selection, allowing users to navigate different functionalities within the system. Additionally, the
interface defaults to a 'Ladies Clothes' selection, indicating a clear starting point for browsing virtual
try-on options. This caters to a common customer demographic while still offering the flexibility to
explore other categories through the menu.

• Prominent 'MENU' Selection: A strategically placed "MENU" button serves as the central
navigation point. This easily identifiable icon ensures users can readily access various
functionalities within the VTO system, fostering a sense of control and exploration.
• Default 'Ladies Clothes' Selection: The interface defaults to displaying "Ladies Clothes,"
catering to a statistically common customer demographic. This jumpstarts the virtual try-on
process without overwhelming users with options. It also acts as a visual cue, instantly conveying
the purpose of the system.
• Flexibility and Personalization: Despite the default selection, the menu offers clear access to
other clothing categories (e.g., Men's, Unisex). This caters to diverse customer preferences and
fosters inclusivity.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 24 of 37
………………………………………………………………………. Virtual Try-On…………

4.2. Menu

Fig.4.2. Virtual Try-On Menu

1.2.1 Click Images:

Icon Description

The 'Click Images' option within the menu allows users to

capture photos of themselves. This feature could be used for
comparing virtual try-on results to their actual appearance or
for creating outfit combinations with their own wardrobe.

1.2.2 Gender Selection:

Icon Description

Allows users to switch between browsing clothes for men,

women, category to ensure a personalized experience.

1.2.3 Size Customization:

Icon Description

Enables users to select their desired size for the virtual try-on
experience, providing a more accurate representation of how
the garment would fit.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 25 of 37
………………………………………………………………………. Virtual Try-On…………

1.2.4 Garment Up/Down:

Icon Description

Provides users with the ability to virtually cycle through

different clothing options within their chosen category,
streamlining the browsing process.

1.2.5 Discard Item:

Icon Description

Allows users to remove garments they no longer wish to try on

virtually, keeping the interface clutter-free and focused on their
current selection.

1.2.6 Sound Toggle (On/Off):

Icon Description

Gives users control over the audio within the VTO system,
allowing them to mute or unmute sounds associated with button
clicks or virtual interactions.

1.2.7 Exit:

Provides a clear exit option for users to terminate the virtual

try-on session and return to the main shopping floor interface.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 26 of 37
………………………………………………………………………. Virtual Try-On…………

4.3 Male Clothes Section

Fig.4.3. Virtual Try-On Male Clothes Section

Selecting a Core Garment:

Clicking on a thumbnail or tapping a category icon opens a detailed view. This view might include:
• Large, high-resolution image: Showcasing the garment in full detail.
• Color and size options: Users can choose from a color palette or a size dropdown menu to ensure
a perfect fit for the virtual avatar.
• Zoom functionality: Allowing users to inspect fabric texture, patterns, and stitching closely.
• 360° view (optional): Providing a complete visual representation of the garment from all angles
(helpful for jackets or detailed shirts).

Accessories (Hat, Tie, Sunglasses, Bag/Suitcase):

• A dedicated section presents various accessories with clear icons or thumbnails.
• Multiple color options: Users can virtually try on different colors of the same hat, tie, or
sunglasses.
• Style variations (optional): For hats, this could include options for baseball caps, fedoras, or
beanies. Sunglasses could offer classic aviators, wayfarers, or sportier styles.
• Size selection: This would be relevant for bags and suitcases, allowing users to choose the
appropriate size for their needs.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 27 of 37
………………………………………………………………………. Virtual Try-On…………

4.4 Female clothes section

Fig.4.4. Virtual Try-On Female clothes section

==================================================================
K. K. Wagh Polytechnic, Nashik Page 28 of 37
………………………………………………………………………. Virtual Try-On…………

Customization Options:
• Each garment category features a selection of thumbnails or icons representing different styles,
colors, and patterns. Users can easily tap or click on these options to virtually dress the avatar in
their chosen item.
• Size Selection: A dedicated size selection panel allows users to choose their desired size for each
garment, ensuring an accurate representation of fit during the virtual try-on experience.

Exploring Garment Categories:

• Lower Wear: Users can browse a variety of skirt lengths, styles (e.g., pencil skirt, maxi skirt),
and materials (e.g., denim, leather). Options for shorts, leggings, and jeans can be presented
similarly.
• One-Piece: This section showcases dresses, jumpsuits, and rompers in various styles (e.g.,
casual, formal, cocktail) and lengths.
• Tops: A wide range of tops like shirts, blouses, sweaters, and jackets can be presented with
options for color, pattern, sleeve length, and neckline.

Accessory Selection:
• The "Accessories" section features a variety of sunglasses and handbags.
• Users can select and virtually try on different sunglasses styles (e.g., aviators, cat-eye, wayfarers)
and colors to complement their chosen outfit.
• Handbag options can be displayed based on size (e.g., clutch, tote, crossbody), style (e.g., formal,
casual), and color. Users might be able to virtually "hold" the bag to see how it scales with the
outfit.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 29 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 5
TESTING OF SOFTWARE APPLICATION

5.1 Unit Testing of Individual Modules

Modules to be Tested (Considering Kinect Integration):

• Kinect Sensor Integration Module:
o Test successful connection and communication between the Kinect sensor and the VTO
system software.
o Verify accurate user body detection and skeletal tracking by the Kinect sensor.
o Ensure proper data transmission of user posture and movement for virtual garment
fitting.
• 3D Garment Rendering Engine (with Kinect Integration):
o Test real-time rendering of 3D garments based on user body data captured by the Kinect
sensor.
o Verify accurate garment positioning and alignment on the virtual user model generated
from Kinect data.
o Ensure smooth animation of virtual garments as users move (e.g., arms raised,
bending).
• Sample Test Cases:
o Simulate various user positions and movements in front of the Kinect sensor to verify
accurate body tracking.
o Test the system's response to different body types and clothing sizes to ensure proper
garment fitting visualization.
o Conduct stress tests with multiple users in front of the Kinect sensor to assess system
performance.
o

5.2 Integration Testing

Modules to be Integrated (Considering Kinect):
• Kinect Sensor Integration Module with User Interface:
o Test seamless data flow between the Kinect sensor, user interface elements (e.g., size
selection), and the 3D rendering engine.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 30 of 37
………………………………………………………………………. Virtual Try-On…………

o Verify that UI elements like size selection dynamically adjust based on user body
dimensions captured by the Kinect sensor.
• VTO System with Inventory Management System (optional):
o Integrate the VTO system with a mock inventory system to simulate real-time stock
checks based on garment selection and user body dimensions.
Sample Test Scenarios:
• Simulate user interaction with the UI (e.g., selecting a garment) and verify that the system
displays the garment virtually on the user model generated from Kinect data.
• Test data exchange between the VTO system and the inventory management system to ensure
stock availability checks consider user body dimensions captured by the Kinect sensor
(optional).

5.3 User Acceptance Testing (UAT) with Kinect Sensor

Participants:
• Target Customers (consider diverse groups):
o Recruit customers with varying body types and clothing sizes to test the accuracy of
virtual garment fitting based on Kinect data.
o Include individuals with different levels of technical proficiency to assess user-
friendliness.
• Store Employees:
o Train store employees on how to use the VTO system with the Kinect sensor to assist
customers effectively.
o Involve them in UAT to test functionalities from an employee perspective (customer
interaction, troubleshooting).
Sample UAT Tasks (Considering Kinect):
• Customers:
o Try on various garments virtually using the Kinect sensor, assessing the accuracy of
virtual fitting based on their body dimensions.
o Provide feedback on the intuitiveness of using the Kinect sensor for virtual try-on.
o Test the responsiveness of the system to different user movements (e.g., raising arms,
bending).

==================================================================
K. K. Wagh Polytechnic, Nashik Page 31 of 37
………………………………………………………………………. Virtual Try-On…………

• Store Employees:
o Simulate customer interactions using the VTO system with the Kinect sensor to guide
virtual try-on experiences.
o Test functionalities like recommending outfits or suggesting size adjustments based on
the user's virtual image generated by the Kinect sensor.
o Identify any potential challenges customers might face using the Kinect sensor and
provide feedback on system ease of use from an employee standpoint.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 32 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 6
COSTING OF PROJECT

- Size of the project: Let's reduce the estimated size to 2,000 lines of code (LOC).
- Effort Multiplier: 1.0 (standard value).
- Labor rates: 500 INR per hour.

Using the Intermediate COCOMO formula:

[ Effort = a times (Size)^b times EAF]

Where:
- (a = 2.8) for Intermediate COCOMO
- (b = 1.20) for Intermediate COCOMO
- (Size) is the estimated size of the project in KLOC (Kilo Lines of Code)
- (EAF) is the Effort Adjustment Factor (effort multiplier)

[ Effort = 2.8 times (2) ^ {1.20} \times 1.0]

[ Effort ≈ 2.8 times 2.297396]

[ Effort ≈ 6.4314]

Now, let's calculate the cost:

If we assume an average of 160 hours per month per person:

[ Total Hours ≈ 6.4314 times 160]

[ Total Hours ≈ 1029.024]

Now, if we multiply this by the labour rate:

[ Total Cost ≈ 1029.024 times 500]

[ Total Cost ≈ 514512

So, the estimated cost of your project using the adjusted COCOMO model is approximately 51,451.20
INR.

The updates in the 3D garments will be subscription based*

==================================================================
K. K. Wagh Polytechnic, Nashik Page 33 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 7
APPLICATION OF VIRTUAL TRY-ON SYSTEM

Virtual try-on technology offers a range of exciting possibilities across various sectors. Here's a
breakdown of its potential applications in different areas:

7.1 Retail Industry:

• Enhanced Customer Experience: Virtual try-on allows customers to experiment with clothes
and accessories in a fun and interactive way, leading to more informed purchases.
• Reduced Return Rates: By providing a realistic idea of fit and style, virtual try-on can
significantly decrease return rates for retailers.
• Increased Sales Conversions: A more engaging shopping experience with virtual try-on can
translate to higher sales conversion rates.
• Inventory Management: Virtual try-on data can inform retailers about popular styles and sizes,
allowing for better inventory management and reduced stockouts.

7.2 Fashion Designers and Brands:

• Improved Design and Development: Virtual try-on can be used to test garment designs on
various body types before production, leading to better fitting and more inclusive clothing lines.
• Streamlined Product Feedback: Designers can gather real-time customer feedback on virtual
try-on experiences, allowing for quicker design iterations and improvements.
• Personalized Marketing Campaigns: Virtual try-on data can be used to personalize marketing
campaigns by recommending products based on customer preferences and body types.

7.3 Online Marketplaces:

• Standardized Product Representation: Virtual try-on can ensure consistent product
representation across different vendors on online marketplaces, improving customer experience
and trust.
• Increased Platform Value: Offering virtual try-on features can make online marketplaces more
competitive and attract more customers.
• Data-Driven Recommendations: Virtual try-on data can be used to recommend complementary
items to customers based on their virtual try-on sessions.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 34 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 8
FUTURE SCOPE OF THE PROJECT

Our virtual try-on system is designed for continuous evolution and innovation. Here's a roadmap
outlining some exciting possibilities for the future:

8.1 Deeper Integration with AI and Machine Learning:

• Enhanced Personalization: Leveraging user data and preferences, the system can personalize
the experience by recommending clothing styles, sizes, and complementary items that are most
likely to suit the user's taste and body type.
• Advanced Fit Prediction: By incorporating sophisticated machine learning algorithms, the
system can analyze user body dimensions and predict the fit of various clothing items with even
greater accuracy.

8.2 Expanding the Virtual Wardrobe:

• Beyond Apparel: The system can be extended to encompass virtual try-on for accessories,
jewelry, cosmetics, and even eyewear, allowing users to experiment with a wider range of
products.
• Customization Options: Users can personalize virtual clothing items by selecting different
colors, patterns, and textures, creating a truly customized virtual try-on experience.

8.3 Integration with Augmented Reality (AR):

• Immersive Try-On: Imagine trying on clothes that appear to drape and move realistically on
your body in real-time! Through advanced AR integration, the virtual try-on experience can
become even more immersive and interactive.
• Spatial Recognition: By incorporating spatial recognition technology, the system can allow
users to virtually try on clothes within their own environment, providing a realistic sense of how
the item would look in their space.

8.4 Advanced Data Analytics:

• Retailer Insights: The data collected from virtual try-on sessions can provide valuable insights
for retailers. This data can be used to optimize inventory management, product development, and
marketing strategies.
• Personalized Recommendations: By analyzing user data and virtual try-on behavior, the
system can generate more personalized product recommendations for individual users.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 35 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 9
CONCLUSION

The proposed virtual try-on system presents a revolutionary approach to fashion shopping. Unlike
traditional methods that rely on static images and size charts, virtual try-on leverages cutting-edge
technology to create a realistic and interactive experience. This empowers users to make informed
purchasing decisions, potentially reducing return rates and boosting confidence.

This innovative solution transcends the limitations of physical stores, offering convenience and
accessibility for customers. Virtual try-on technology allows customers to digitally try on clothing and
accessories, providing a realistic and interactive shopping experience. By eliminating the need for
physical try-ons, it saves time and enhances convenience for shoppers. With its potential to increase
customer engagement and boost online sales, virtual try-on is a valuable tool for the retail industry.

Furthermore, virtual try-on unlocks a wealth of data for retailers, enabling them to optimize inventory
management, product development, and marketing strategies. As virtual try-on technology matures, it
has the potential to reshape the online fashion retail landscape, creating a win-win situation for both
consumers and businesses.

==================================================================
K. K. Wagh Polytechnic, Nashik Page 36 of 37
………………………………………………………………………. Virtual Try-On…………

Chapter 10
REFERENCES

Academic References:
• Isıkdogan, F., & Kara, G. (2012). A real time virtual dressing room application using kinect.
CMPE537 Computer Vision Course Project.
• Korszun, H. A. (1997). U.S. Patent No. 5,680,528. U.S. Patent and Trademark Office.

Web References:
• https://openaccess.thecvf.com/content/WACV2022/papers/Majithia_Robust_3D_Garment_Digi
tization_From_Monocular_2D_Images_for_3D_WACV_2022_paper.pdf
• https://paperswithcode.com/paper/viton-an-image-based-virtual-try-on-network
• https://paperswithcode.com/task/object-detection

==================================================================
K. K. Wagh Polytechnic, Nashik Page 37 of 37