I'm glad to announce a couple of updates of PyPTV and OpenPTV:
- openptv segmentation.c now has an artificial single target added if no targets in the image is found. this workaround protects against segfaults when we play with the object detection parameters
- PyPTV has a new detection_gui.py that allows to load an image and create detection parameters dynamically, i.e. sliding threshold, min-max pixel size, etc.
- PyPTV works also with black-on-white objects, there is a checkbox "inverse" that automatically inverts the image. it works for the manual processing as well as for the sequence
OpenPTV can be installed using:
2. cd openptv/py_bind
3. python setup.py prepare && python setup.py install
Then PyPTV can be installed by pip directly from git
PyPTV does not require the update, but it's safer to play dynamically with the detection with this new workaround trick.
Good luck with 3D-PTV
Alex