goldTracking

Algorithm to track gold nanoparticle self assembly with known number of particles

mainTracking

The main code to be used is called mainTracking.m. It can currently track up to 8 particles and is expecting fluorescence/darkfield type data (bright foreground). It can currently process .mp4 and .ome.tif as well as entire folder of the two (processing every file in the folder. It is expected to be given a folder in which a single file is OR a folder in which subfolder each containing single files are.

To run, the user needs to provide several input:

1. Path to the folder to analyze (UI), this you will be requested to provide upon running the code
2. delta - radius around the particles to be picked in pixel. This is to reduce the processing time, smaller is better but you need to be sure to not crop smaller than where you particles are moving
3. number of particles - This code is made to analyze and always find particles, but it needs to receive the correct number of particles preamptively
4. width - width to be used for the gaussian fitting 3 for 200 nm beads and 95nm pixel for instance. 0 Will let the width a free parameter
5. pxSize - pixel size in nm
6. minDist - minimum distance expected between particles in pixels
7. scaleBar - size of the scale bar for gif in um
8. tail - number of frame of the tracked trace to be plotted on the gif movie for every single frame
9. frameRate - frameRate of the movie
10. info.type - 'normal' or 'transmission' transmission will make the code revert the colorscale to create a bright foreground image
11. OutputFolder - name of the folder to be created to store the results. The folder will be created in the path of the file currently analyzed

Once all the parameter are set, you can run the code which will ask you to navigate to where your data is stored. The code will output a figure with the localization and a gif with the localization on the movie played.

The fitting is done via Gaussian fitting, multiple gaussian are fitted at the same time with centers placed wherever the initial detection positioned the particles.

The tracking is done via munkres minimization of the squared distance.