This repo contains a couple of scripts to convert video (.avi and .mpg) to sequences of frames for use with AI and machine learning methods.
The input is a folder with videos to be converted.
The output is a folder structure based on a Patient > Sequence > frame pattern. Each patient has its own folder, and the folder may contain one or multiple folders with video sequences. A sequence is a set of image frames numbered from 1 to N.
There is one simple and one more "advanced" script. Adjust the INPUT_DATA_PATH and OUTPUT_PATH variables and run the script to generate folders and image frames.
Of course, feel free to adjust the script according to your needs!