This repository contains the code for the journal article Spatiotemporal Gait Measurement with a Side-View Depth Sensor Using Human Joint Proposals, which is available on IEEE Xplore.

Requirements

Python version 3.7 or higher is required.

Some scripts for the results and figures make use of LaTeX for formatting. You can either install LaTeX or edit the code so it is not needed.

Setup

Create a virtual environment and activate it.

$ mkdir ~/.virtualenvs/
$ python -m venv ~/.virtualenvs/depth_gait_env

$ source ~/.virtualenvs/depth_gait_env/bin/activate

Once the environment is activated, install the required Python packages.

$ pip install -r requirements/base.txt

Usage

Method

The following scripts constitute the method of the paper.

Estimate the lengths of body links on each trial:

$ python -m scripts.main.estimate_lengths

Select the best proposals for head and foot positions:

$ python -m scripts.main.select_proposals

Cluster the frames of the walking trials to determine the walking passes:

$ python -m scripts.main.label_passes

Calculate gait parameters from the selected head and foot positions:

$ python -m scripts.main.calc_gait_params

For convenience, all of these scripts can be run at once:

$ python -m scripts.main.run_all_main

Results

Run all of the results:

$ python -m scripts.results.run_all_results

Figures

Generate all of the figures used in the paper:

$ python -m scripts.figures.run_all_figures