uu-cellbiology / dom_utrecht Goto Github PK

check the precision of results storage
also for export/import

Save particle table with averaged localizations of particles in subsequent frames

MLE fitter
Localization error as https://www.ncbi.nlm.nih.gov/pubmed/24302478 equation 7 (and 5)
Derivatives for fitting: https://www.osapublishing.org/oe/abstract.cfm?uri=oe-24-5-4996 equations 13 and 14
Derivatives of mu_k = N/(2pisigma^2)*exp(-[(x_k-x_0)^2 + (y_k-y-0)^2) / (2 * sigma^2) + b
can be derived by hand

Hi @ekatrukha
Many high/super resolution methods can output 2, 3, or even 4 colour channel images.
Procedure could be repeated for further colour channels, but then need a way to know which colour channel image is which...
... since the order of the image colour channels might not always be the same in different imaging experiments from the same optical setup? Today RGB, tomorrow BGR

perhaps use meta data from image import to determine and label illumination or emission channel by wavelength, eg 488 nm or "colour" - blue, green, red, far red, cyan, yellow, orange, magenta, etc.

Or just have channels labelled REF, 2, 3, 4 etc, and let the user worry about getting them in the right order by making sure the calibration image and the image to correct have colour channels in the same order.

Would need to support composite images?

by reconstructing only user selected small area
(no need for whole image reconstruction, actually)

PALMsiever: https://github.com/PALMsiever/palm-siever

ViSP software

point cloud loader www.cake23.de

Convolve with eliptic Mexican hat

improve image histogram generation for the fitting