1

Background

I have multiple, synchronized cameras with partial overlapping views. I already have object tracking applied to the objects in the videos I get from these cameras. Now I want to have a tool that shows me these videos, including the tracks, and enables me to say that a certain set of tracks in different videos belong to the same object.

Basically I am searching for a free, accessible and more developed alternative for the following annotation tool:

enter image description here Code: https://github.com/cvlab-epfl/multicam-gt

Note that the above tool seems to suit my needs, but I wonder if there is not a more professional tool available.

For my needs the CVAT annotation tool is almost perfect, except for the fact that only a single view can be annotated at the same time, thus making it hard to annotate in a multi camera setup.

Must haves

  • Support for bounding box annotation of multiple camera views add the same time.
  • Free of charge
  • Runnable on my own pc, i.e. not in the cloud

Nice to haves

  • Works with camera calibration, i.e. adjusting the location of a bounding box in camera 1 changes the bounding box locations in different camera views once these bounding boxes have been linked.
  • Works with masks instead of bounding boxes
  • Multi user support
  • Deep learning model support
  • Python code base
  • Runs under Linux

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.