A real time virtual dressing room system which consists of multiple tasks including detection of the user’s face from video stream, alignment of models, approximating the position of torso and lower body based on face detection and resizing input dress images and dress up using Image Processing.
MICROSOFT among its various innovative products gave the world, the KINECT. Along with it they showed us some possible applications in which it could be used, and among it was the Virtual Dressing Room, which was based along a similar problem statement. But the model was applicable only for showrooms which could afford and actually own the Kinect setup. We thought of designing a product which was along the same problem statement but without the technical and hardware barriers. Our model is applicable for any user who has a basic RGB camera on his device. Thus it can be extended to online shopping users and smaller businesses who wouldn’t have opted for a costly Kinect.
Run the flasktry.py file and go to localhost. Once on the main page, you can check out various shirts and pants from the navbar. Predict will show you the result with the selected shirt and pant.
pip3 install opencv-python
pip3 install flask
Webcam/RGB Camera
- OpenCV
- Flask