# yolo_archery_tools

**Repository Path**: xep0268/yolo_archery_tools

## Basic Information

- **Project Name**: yolo_archery_tools
- **Description**: Simple utilities for analyzing archery videos with YOLO
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-01-11
- **Last Updated**: 2025-01-12

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Simple utilities for analyzing archery videos with YOLO

This is a simple Python package that providing some (very simple) tools for using [ultralytics](https://github.com/ultralytics/ultralytics) to analyze archery videos.

## Install

This package could be installed using standard Python package managers. To avoid version issues of [pytorch](https://pytorch.org/), it is suggested to install `pytorch` in advance with the command line helper on its [website](https://pytorch.org/).

For `pip`, simply use the following command at the root directory of current project is enough:

```shell
pip install .
```

## Usage

This package provides some command line tools that could be simply called from the shell.

### Generating keypoint data

To generate the keypoint data (i.e., the keypoint coordinates of each frame in the video), use the command line like this:

```shell
python3 -m yolo_archery_tools.run_and_save -m <model_name> -s <source_video_path> [-o <output_file_name>]
```

where, `<model_name>` is the YOLO pre-trained model name, which could be found from ultralytics document: [Pose - Ultralytics YOLO Docs](https://docs.ultralytics.com/tasks/pose/#models); `<source_video_path>` is the path (in local file system) of the video to be analyzed, and `<output_file_name>` is the keypoint data file to be written (in CSV format).

### Analyzing with workflow

Use the command line tool: 

```shell
python3 -m yolo_archery_tools.workflow <workflow_file_name>
```

The workflow specification file should be written in [TOML](https://toml.io) format. Example workflow files are provided in the `examples` folder.