nplocate
is a custom script I wrote to locate very tiny particles from a confocal image. These images often suffered from extreme influences of the PSF, even after very detailed and completed deconvolution procedures.
To squeeze a bit more information out of these highly distorted data, I wrote this code to effectly "fit" the entire 3D image. This is done in a quite sloopy way. For a perfect fit, please take a look at the very well crafted peri project.
This is not a fully functional particle tracking package like trackpy or colloids or peri. Instead, think of nplocate
as an extension of current tracking packages.
The logic behind the code is quite simple. The arguments are,
- It is easy to find some particles, even in a highly distorted image.
- If we know the locations of some particles ({r}), we can measure their average shape (S).
- With {r} and S, we can simulate a "fake image"
- We can find previously unfound particles in the difference between the real image and fake image.
- The more particles we have, the merrier.
The simplest way is
pip install nplocate
You can also download this repository, and use the following command to install the code
pip install .
There are some notebooks in the folder example
that introduced how to use this package, along with trackpy.
Just tell people you used trackpy