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.
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.
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
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
Run all of the results:
$ python -m scripts.results.run_all_results
Generate all of the figures used in the paper:
$ python -m scripts.figures.run_all_figures