Quickstart Guide#

This guide shows the simplest end-to-end usage for available modules.

Tracking#

Create local config files (optional but recommended if you want to tweak defaults):

bee config init

  1. Run the Tracking pipeline on a video:

bee track run-pipeline --input /path/to/video.mp4 --output /path/to/output_dir

  1. Inspect the outputs (CSV files) in the output directory:

  • *_yolo_tracking_raw.csv: raw detections

  • *_yolo_tracking_fixed_ids.csv: detections with stable IDs

  • *_yolo_tracking_fixed_ids_velocity.csv: flow and speed metrics

  1. Visualise tracks on the source video:

bee track visualise-tracking --video /path/to/video.mp4 --csv /path/to/fixed.csv \
  --out /path/to/annotated.mp4

Batch processing#

Batch processing allows you to run the pipeline across all videos in a directory tree.

bee track batch-process --input-dir /path/to/videos

This will create a new directory called tracking_outputs in the current working directory, with subdirectories for each video containing the outputs.

To specify a different output directory name:

bee track batch-process --input-dir /path/to/videos --output-dir-name tracking_outputs

To filter videos by name (e.g. only process videos containing “hiveA” in the filename):

bee track batch-process --input-dir /path/to/videos --filter hiveA
On this page