WaggleDanceTracker

Data Science MSc - City, University of London Final Project

Description

The code should be run in the following order:

1. DanceDetector.py
2. DetectionCleaning.py
3. WaggleRunTracker.py
4. ClusterRuns.py
5. SaveDances.py

DanceDetector.py

• Requires a static video of an observation hive with any number of bees performing the waggle dance.
• Takes the video and detects 'waggle-like' activity based on contour detection.
• Clusters the activity based on density to isolate waggle runs.

DetectionCleaning.py

• Removes duplicate points (two coordinates at the same frame in the same cluster).

WaggleRunTracker.py

• Using the detection dataset as a skeleton, tracks the bee throughout the waggle run.
• Saves data on orientation, waggle frequency, spatial coordinates for each cluster.

ClusterRuns.py

• Clusters each waggle run into a waggle dance based on a forward search (within 5s of the previous run ending).
• Clustered based on proximity.

SaveDances.py

• Saves each waggle dance as a separate cropped video.
• The bee is highlighted during the waggle run as a visual aid.

Paper: https://www.biorxiv.org/content/10.1101/2020.11.21.354019v1