Optical Finger Tracking

This is meant to use a video file and a log file to determine what finger was used in each key press.

Requirements

You need at least python 3.6 and opencv. Install both with the package manager for your system (recommended) or (if this is not possible) install python if you don't already have it and then install opencv-python with pip (python -m pip opencv-python), this should install numpy as well.

Recording the video

Record the video in the avi format! The resolution doesn't really matter but shouldn't be less than 640x480.

Do not move the camera after starting the recording! If it happens anyway: try to get a clear image (i.e. clear view of the keyboard) as fast as possible to allow for later corrections.

Placing the markers

The fingers should have the following colors:

Finger	Color
thumb	turquoise
index finger	red
middle finger	blue
ring finger	green
pinky	violet

The marker should be in a position that it is in the middle of the finger when viewed from the camera.

Settings

There are some settings available. Have a look at settings.py to learn what those are and how to use them.

Hotkeys

• q to quit
• _ (space) to pause/unpause and end the setup

Running the program

Call python3.6 <path to main.py> <path to the video> in the console. The paths are relative to the folder you're in when executing the call. You only need main.py and settings.py to run the program. You can copy those to whatever location you like.