1. DIGiTR - Digital image Trial Room
Adi Shankara Institute of Engineering Technology, Kalady
Department of Computer Science & Engg
Under
APJ Abdul Kalam Technological University
Name & Role of Project Members :
Akhila Asok (Backend Developer and OpenCV module)
Aparna A V (Frontend Developer and OpenCV module)
Reg no :
ASI17CS007
ASI17CS015
Guide Name : Asso. Prof T Sobha
2. TABLE OF CONTENTS
❑ Objective
❑ Problem Statement
❑ Design Diagrams
❑ Components [H/W & S/W]
❑ Methodology /Algorithm/Architecture of the project
❑ Features of the the Project
❑ Status of Project – Implementation
❑ Code Sample / Product Video / Screenshots
❑ Details Of Publication
02-06-2021
DIGiTR
2
3. ● Initiative to help improve the conditions in the textile sector due to this Covid
pandemic and to boost sales in the sector.
● As the lockdown eased, most of the clothing stores opened with new practices
such as ‘no trial’ and ‘no exchange’.
● DIGiTR is a smart mirror that can help customers digitally try on a dress. It
lets the customers
● Views a 3D image of themselves to see which dress fits them better.
OBJECTIVE
02-06-2021
DIGiTR
3
4. The COVID-19 pandemic has affected our society very much in a lot of ways.
It has changed the way we do things, work, and the way we shop. And it has also
affected many businesses in problematic ways. One of the evident businesses that
were affected by this pandemic is the Textile businesses. Because of the current
scenario, it became indispensable to avoid direct contact with people and therefore
making it difficult to try on dresses and clothes to see if they are suited for you.
Recently there has been a lot of concerns surrounding catching Covid-19 in trial
rooms. Trial rooms in retail outlets are usually poorly ventilated. With the Covid
pandemic spreading wildly across the globe, shoppers are wary of entering stores
that can get crowded rapidly. And with social distancing rules, shops have also
been forced to seal off their trial rooms temporarily.
PROBLEM STATEMENT
02-06-2021
DIGiTR
4
8. COMPONENTS [H/W & S/W]
Hardware Components :
● Raspberry Pi 4 Model B
● Camera
● Intel Neural Compute Stick 2
Software Components :
● Fast API:
● MongoDB
● Angular 2.9
● Open CV
● Numpy
● Uvicorn
02-06-2021 DIGiTR 8
9. The system includes very few hardware components which are Raspberry Pi 4, the brain
of the module, a 1080p Pi cam and a neural compute stick 2 by intel which handles image
processing functions with an improved capability. The majority of the project lies on its
software end, where one uses the computer vision fundamentals to bring this project to life.
OpenCV expanded as Open Source computer vision and machine learning software
library, including programming functions mainly aimed at solving real-time computer vision
problems. It is by using this library that we have developed our ML model that is used to
produce a "try-on" feature. The working of the project from the vantage point of the user can
be explained in a few simple steps:
Methodology /Algorithm/Architecture of the project
02-06-2021 DIGiTR 9
10. 1. The product is designed to constitute a large plane 2-way mirror with a camera mounted on
top. The mirror constitutes a 20-inch display on the side, which the user can use to log into
the e-commerce application.
2. The customer selects a dress from an array of collections presented on the 2-way mirror .
3. On selecting their desired dress, the camera module is activated, which feeds the live
image of the customer.
4. The customer is then asked to step back to see themselves wearing the dress of their choice
on the screen of the 2-way mirror. The display even recommends the appropriate size of
the dress that fits the use.
The product also includes a retailer version of the app, which can be used by staff members
of shops to add more collections to their inventory.
Methodology /Algorithm/Architecture of the project
02-06-2021 DIGiTR 10
11. ● Shopkeeper Login: This function helps shopkeepers to log in to the product
using the user id and password given when they register to the product.
● Adding Variable Products: This function helps staff to add new clothing to the
product.
● Removing Products: This function helps staff remove unwanted clothing from the
product.
● Selection of Desired Dresses: A random dress is displayed on the screen where
the customer can scan select their desired dress.
● Current Stock Updation: This function allows the user to see whether the
desired dresses they selected have been recently bought and are in stock or out
of stock.
● Customer Try on Function: This function helps the customer to try on virtually
the decided clothing whether by uploading photo or by video
● Add To Cart: This function lets the customer select the desired dress they
selected to buy.
❑ Features of the the Project
02-06-2021 DIGiTR 11