Skip to content

SheffieldMLtracking/btretrodetect

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

89 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

retrodetect

  • This new version (completely rewritten) assumes all the photos are flash photos.
  • Also handles colour camera!

install

[to add]

usage

One passes a path and the tool will recursively search through the subdirectories, finding all the images, sorting them (within that folder) and applying the retrodetect algorithm.

Example: btretrodetect ~/Documents/Research/rsync_bee/test/beephotos/2023-06-29/sessionA/setA/cam5/02D49670796/ --after 10:32:29 --before 10:33:29 --threshold -10


Runs the retoreflector detection algorithm

positional arguments:
  imgpath               Path to images (it will recursively search for images in these paths)

options:
  -h, --help            show this help message and exit
  --after AFTER         Only process images that were created after this time HH:MM:SS
  --before BEFORE       Only process images that were created before this time HH:MM:SS
  --refreshcache        Whether to refresh the cache
  --threshold THRESHOLD
                        Threshold of score before adding to data
  --sourcename SOURCENAME
                        The name to give this source of labels (default:retrodetect)```

workflow

  1. Use btqviewer . to label bees (this makes a folder, btviewer, in the places where you do this).
  2. Use btretrodetect-train . in this location (ideally across a session or set, i.e. from multiple cameras). This by default replaces the current model with a new one (saves the old one with a different name).
  3. Try out btretrodetect on a new dataset.
  4. Load the new dataset with btqviewer to see how it does.

About

Detect retroreflectors in images

Resources

License

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Jupyter Notebook 94.7%
  • Python 5.3%