Skip to content

Latest commit

 

History

123 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Pole Vault Video Analyzer

Table of contents

Technologies

Project is created with:

  • OpenCV version: 4.5.1
  • OpenCV_contrib modules version: 4.5.1
  • ffmpeg version: 4.3.1 (older versions should work as well)
  • gtk+ version: 2.24.32_3 (older versions should work as well)
  • cmake version: 3.1 or later
  • python version: 3.8.2 (older versions should work as well)
  • matplotlib version: 3.3.2 (older versions should work as well)
  • argparse version: 1.1 (older versions should work as well)

To install OpenCV and OpenCV_contrib, run script install_opencv.sh. The library will be installed to ~/.local.

Note that ffmpeg and gtk+ must be installed before OpenCV is installed.

Correct functionality requires trained models for body parts detections. You can download them running script models.sh.

Note that models.sh and install_opencv.sh uses wget package.

Setup

To run this project, install it by running script install.sh.

This creates executable file build/PVAnalyzer.

Usage

Program can be run in two modes:

  • Video analyzer
  • Model analyzer

After video or model is analyzed, you will be able to view all video frames with detections drawed in them. Navigate through frames using left and right arrows, parameters for each frame will be written to console. Note that only parameters valid for viewed frame will be written, for example velocity losses computed during takeoff will be written only in frame showing sthlete's takeoff. You can turn detections drawings off/on by pressing space bar. To end viewer, pres escape key.

To show help, use single argument --help.

Video analyzer

To analyze videos, pass videos' paths as command line arguments:

$ ./build/PVAnalyzer vid1 vid2 vid3

You can specify flag -f before first video to perform automatic athete's detection.

Command above analyzes videos vid1, vid2 and vid3. You can pas any number of videos to be analyzed sequentially. Video analyzer mode creates output files in folder outputs/X/, where X is name of analyzed video (without path and extension).

Model analyzer

To analyze saved models, pass models' paths as command line arguments:

$ ./build/PVAnalyzer -m mod1 mod2 mod3

Command above analyzes models mod1, mod2 and mod3. You can pas any number of models to be analyzed sequentially. Model analyzer mode does not create any output files.

Plotting

Output parameters can be ploted as graphs using plot_params.py script.

To plot parameters saved to file output.csv, run the following command:

$ python3 plot_params.py --file output.csv

This script will show graphs of all parameters stored in file output.csv.

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages