Virtual Dressing Room

Our approach is to detect the user by using video streaming, first of all, the problem is to detect the user and the body parts detection is also an important step. This type of problem can be managed by simple software like OpenCV or the Visual Studio. Since the extraction of user image in order to

Project Title

Virtual Dressing Room

Project Area of Specialization

Augmented and Virtual Reality

Project Summary

Our approach is to detect the user by using video streaming, first of all, the problem is to detect the user and the body parts detection is also an important step. This type of problem can be managed by simple software like OpenCV or the Visual Studio. Since the extraction of user image in order to create the augmented reality environment by isolating the user area from the video streaming and the superimposing the clothes onto the user.

Project Objectives

We know that trying clothes in the trial room is a time-consuming process and may cause a crowd in the store. Unlike books, groceries, and other stuff which are easy to select, clothes need lots of choices in color, size, and quality clothes. Therefore, some people take time in selecting the right product but some don’t bother to try every apparel and just purchase an item without satisfaction. Furthermore, many stores are anxious about theft or clothes damage in the trial room. This project can overcome all these and many more issues. This product is beneficial for both store management and customers.

Project Implementation Method

The image of the user is taken by the webcam and then by using the image as an input it is transferred to the software module. Further, the techniques like body and face detection are being done so that the video is scanned for the presence of human faces. Lucas-Kanade (LK) tracking is done as it is used for layered motion, mosaic construction, and face coding after this these images are then masked to protect some areas of the image from being altered as changes will be made to the rest of the image. Segmentation is being implemented by the Viola-Jones and fitting is implemented by the Graph Cut algorithm Finally the masked image is superimposed by the image of the dress of the user’s choice and the user is now able to see his/her image wearing that chosen dress and can buy product.

Benefits of the Project

- Save time for clothes trial.
- Easy selection of color and size
- Saves clothes damage in store
- Reduce clothes theft
- Personal Hygiene: Saves people from many diseases like skin disease, eczema, etc.
- Easy update of every new arrival (through the remote database)
and more other benefits...

Technical Details of Final Deliverable

Here we will use OpenCV which will detect which have special functions that are used for capturing images and video. It is capable of reading a live video feed from a cam and then for further processing, it breaks the image/video into frames so that it would be advantageous for processing in further stages. The background is removed by blending the RGBA image that we do normalization at pixel level by setting the alpha channel to zero is the pixel doesn’t lie on the user. The video or images are not being saved in this application as this will require more space and it is ethically wrong to save anyone’s image/video without their permission.

? Face Detection: In this step, the face is being detected by the user, the output of this step is in the form of patches containing each face in the input image. For this purpose, the VIOLA JONES algorithm is used, it is the first object detection framework that provides detection of objects in real-time Input user video Face detection User body detection segmentation Mask image Translation & Rotation Scaling Fitting Final output video and this algo was proposed in 2001 by Paul Viola and Michael Jones, this algo is carry out in OpenCV as cvHarrDetectObjects ().

? Segmentation: This segmentation is done to simplify the image as it is easier to analyze the image in a segmented way. It is the process of assigning a label to every pixel in an image. Each label has some visual characteristics, so this is computed in terms of color, texture or intensity.

? Fitting: The parameters that are being used in it are following:
• Scaling: The measuring of distance of object from the screen is done by scaling.
• Translation: The detection of side distance from right and left corner is done by translation.
• Rotation: The rotation angle of the object is done by rotation

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Media

Other Industries

Core Technology

Augmented & Virtual Reality

Other Technologies

Sustainable Development Goals

Good Health and Well-Being for People, Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com