FaceFilter

Face Filter is a program that places select filters onto faces that are detected in the given image or the live feed.

Motivation

We were inspired by popular Instagram and Snapchat filters and wanted to see if we could create a makeshift version of them in only two days for the 2020 Winter Hacklympics Hackathon hosted by MLH.

Screenshots

Features

Face Filter uses live facial detection and image processing operations and places a filter onto each face detected by the device camera. Through a number of calculations, it determines the optimal placement, scale, and angle of the filter mask dependent on each individual face.

Installation

Windows

Use the package manager pip to install cv2, numpy 1.19.13, and scipy in command prompt.

pip install opencv-contrib-python
pip uninstall numpy
pip install numpy==1.19.3
pip install scipy

Mac OS

coming soon...

Usage

Image Filter

1. Open the filter.py file to edit
2. Change these lines to include the name of the image and the respective filter that you want to use.

#read images
img = cv2.imread('assets/image_name.jpg')
fltr = cv2.imread('assets/filter_name.png')

3. Run the filter.py file
4. Press any keyboard button to exit

Live Video Filter

1. Open the live_filter.py file to edit
2. Change this line to include the name of the filter that you want to use.

#read images
fltr = cv2.imread('assets/filter_name.png')

3. Run the live_filter.py file
4. Press any keyboard button to exit

Credits

Sooren Ghodsi, Ethan Li, Seth Eshraghi, Joseph Gong, and Mitchell Kireger

License

MIT © Sooren Ghodsi