This repository contains code to correct for shifts and chromatic aberrations between wide field microscopy channels. The implementation is based on a paper by Michal Kozubek, although any bugs or misconceptions are likely to be mine.

Typically these functions are used on calibration beads to set up a mapping between a reference channel (pick the best one) and the other channels.

These files are slightly out of context here as they are part of a bigger toolbox (DOTTER) which also contains a GUI.

To use these files with MATLAB, it might be a good idea to add the corresponding folder to the path, i.e.

addpath('~/where/i/keep/df_cc/')

Note, there is no code to extract dots from images here, and for any practical usage, the input should be fitted coordinates. For exploring these functions, a good starting point might be df_cc_demo1.m.

• df_cc -- Cluster dots to find correspondences, and find correction. Uses the following functions:
• df_cc_cluster -- find correspondences between points.
• df_cc_create -- find the transformation between two or more channels.
• df_cc_view -- metrics about the correction results.

Note: the .cc files used to store data are just normal .mat files that can be loaded with the -mat switch to load.

• df_cc_apply_dots -- transform dots to match a reference channels

• df_cc_apply_image -- correct a single image
• df_cc_apply_image_folder -- correct all images in a folder

• df_cc_demo1 -- demo using a constant shift / dot correction.
• df_cc_ut -- some unit tests

These files are used by the other scripts:

• df_cc_eudist euclidean distance
• df_cc_poly2mat -- create matrix for finding coefficients.

• df_cc_apply_m -- Correct M.dots in a DOTTER .NM file.
• df_cc_apply_n -- Correct all N.userDots in a DOTTER .NM file.

• Use a polynomial also in the axial direction?
• Add some more demos and demo images.