GaitScope

Upload a walking recording from MATLAB (.mat), Physics Toolbox (.csv) or phyphox (.zip), or record one with your phone, and compare step detectors on it: Coza’s peak detector from BME 598/494 (Dr. Aurel Coza), its modified version, and textbook methods.

GitHub

No recording loaded

Pick a file to see its steps. Everything runs in your browser; the file never leaves your computer.

Try a demo walk

To record a walk with the phone’s own sensors, open this page on your phone.

MATLAB .mat files saved by MATLAB or Octave (v5 to v7, and v7.3, which loads a 41 KB reader the first time), like the course’s Walking.mat. Any variable name works. Samples should run down the rows; a time column first is optional.

Physics Toolbox .csv exports from the G-Force Meter, Linear Accelerometer, Gyroscope or Multi Record, with elapsed or clock time, comma or semicolon separated.

phyphox .zip exports (Export data → CSV) of Acceleration with g or without g, as phyphox saves them, or the Raw Data.csv inside. The phone, sensor and start time from its meta folder are listed with the checks.

Other CSV files with numbers in columns, with or without a header row.

GaitScope JSON exports reopen an analysis: the signal, every setting and the notes.

Record a walk on a phone: the page reads the motion sensors itself (about 60 Hz), with no app to install. The recording loads like a file and can be downloaded as CSV, with the steps you counted and where the phone was.