Skip to content

KonKob/Catwalk

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

44 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This repository contains Code, that was used to analyse 2D-positional Data (Walkthrough.ipynb, Analysis). It takes DLC data derived from animal videos, a camera calibration based on a checkerboard video (see calibrate.py), metadata describing the ROI around the area of the experimental paradigm. Converting the positional data to the reference space and data smoothening is then performed. Step Detection and parameter calculation (angles, distances, areas between bodyparts) results in single values reflecting the median of the parameters over all steps or in the psth of parameters over all steps (first and last step of each recording excluded) with step length normalized to 15 frames.

The Code to select the ROI is based on Code from @DSegebarth and @MSchellenberger and can be found under this link: https://github.com/DSegebarth/BSc_MS/blob/32c940778c648770f63776760033b404b9a00aa5/GUI_annotate_maze_corners.ipynb.

For filtering and interpolation of the tracking data Code from @DSegebarth was copied into this analysis. The original Code can be found under this link: https://github.com/Defense-Circuits-Lab/Gait_Analysis/blob/8510de88e42069259981654f7adc8315d282527d/gait3d/gaitanalysis_with_top_cam.py.

Step detection is based on Code from @ELGarulli: https://github.com/WengerLab/neurokin/blob/d077050bc8760abfe24928b53ad6383e96f59902/utils/kinematics/event_detection.py

For markerless position tracking of the animals, we used DeepLabCut (v2.1.8) with a resnet-50 model.

Code style: black

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published