This script is designed to convert generated annotations from the AVAT (https://github.com/AIFARMS/AVAT) into image format.
This project uses python3, pip and venv. Please install them based off of your system requirements.
Setup virtual env:
python3 -m venv vir_env
Setup terminal to virtual environment:
source vir_env/bin/activate
Install dependencies:
pip3 install -r requirements.txt
There are a few inputs needed to the app to function, you will need a video file and the annotation file generated by AVAT.
python3 -i <INPUT FILE PATH> -s <FRAME SKIP VALUE> -a <ANNOTATION FILE PATH>
A sample input would be:
python3 main.py -i /home/test_user/Desktop/video.mp4 -s 15 -a /tmp/generated_annotations-1.json
- Support for segmentation annotations
- More customizable options for output files
- Custom post processing scripts