Skip to content

This is a photodiode based system utilizing a random forrest classifier to recognize hand gestures held above the array.

Notifications You must be signed in to change notification settings

saehuihwang/hand_gesture_recognition

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Hand Gesture Recognition

This repo contains the code necessary to use Python's sklearn library to recognize hand gestures based on input from an array of photodiodes.

Setup Outline

Below are the necessities to get gesture recognition working

  • 6x6 Array of photodiodes
    • Other sizes will work but will require significant rewrites to Arduino code - Python code should be fine
  • Each diode is connected via a transistor to a TIA (trans impedance amplifer)
    • The output of this TIA is connected to an analog input pin on the Arduino
    • The base/gate of each transistor is connected to a GPIO pin on the Arduino
      • This could also be done with less pins using a MUX
  • Lighting on the diodes is controlled such that your hands cast well defined shadows on the diodes
  • +5V supplied to photodiodes
  • If using an OpAmp in your TIA, +5V and -5V supply will be required

LED Array Output

This repo also includes code for outputting the results of the gesture recognition onto a NeoPixel LED array. While this is optional additional setup is required.

  • LED array Din is connected to a PWM pin on the Arduino
  • +5V supply to LED array

As an additional note, the grounds for the TIA, Arduino, and LED array must all be connected.

Worked with Josh Anderson () on this project.

About

This is a photodiode based system utilizing a random forrest classifier to recognize hand gestures held above the array.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published