This is a work-in-progress pipeline to extract frames from a video using opencv then apply a Segment Anything Model to each frame to identify multiple animals. Ultimately, segmented frames will be combined and segments tracked to track multiple animals moving through frames of videos.
- A model from Segment Anything, placed in the 'Model' folder in your directory.
- Segment Anything version found here: link to be able to threshold max and min size of segments.