GAIT BASED AUTHENTICATION SYSTEM

About Project

This project aims to authenticate people based on gait - the way of one’s walking, we initially propose to collect data from android operated smartphone through accelerometer and upon analyzation we can identify the person through his/her way of walking.

Please help us populate data by installing app below, your help is much needed and appreciated. All volunteers will be included in project contributors list.

Already done setup and logged data - Upload it here

Accelerometer Apps

Consider the following link to download and install apps for data logging on your devices, please put down phone in your pocket at appropriate position whenever you're walking naturally - it may be jogging, normal walk, taking stairs.

Remember! More data we've more the perfect precision will be.

Install Recommended App/s [Acceleration Log] : Accelerometer Apps - Google Drive

How to log data from accelerometer using android app listed above

  1. Open App Accelerometer Log


  2. Accelerometer data will start displaying in form of graph


  3. Put your phone instantaneously in pocket in position described :


  4. Start walking naturally - slow, normal or fast
  5. After finishing walk touch options and tap Stop


  6. After finishing walk touch options and tap Save Data



  7. Name the file in conventions mentioned below
  8. Note : This app dosen't works in background, so you've to keep app open during capture, thanks for your understanding.

How to name your accelerometer log files

  1. Your Name initials - vkku for Vivek, vtsl for Vatsal and alike.
  2. Name your pace - factor of : Fast, Normal, Slow
  3. Name your walking surface - factor of : Smooth, Rough
  4. Optional(Recommended) : give a initial of place you recorded data at - example : MDA, MIT etc.

Now, your file name should look like : vkkuNormalSmooth or vtslFastRoughMDA (Recommended).

Done Logging help us by submitting : Upload Here

Fork me on GitHub