Scripts

MI registration

The scripts to evaluate the Mutual Information (MI) registration are in the MI_registration folder.

Compute SIFT

Description: Computes the SIFT features in two sets of images and finds the pairwise transformations between both sets of images. The images sets are sorted alphabetically. The results are exported in a csv file containing the rigid matrix transformation.

Requirements: Fiji must be installed.

Example:

$ fiji --ij2 --run scripts/compute_sift.py 'pathA="/path/*_A.png",pathB="/path/*_B.png",result="result.csv"'

Convert N channels (e.g. RGB) images to grayscale (N-d to 1-d)

Description: Converts multichannel images, for instance RGB images, to grayscale. The transformation is done by computing the principal components of each image per pixel. The new images keep the same name and are exported in the output folder. The output folder is created if it does not exist. The images sets are sorted alphabetically.

Requirements: requirements.txt contains the needed python packages.

Example:

$ python3 scripts/nc2gray.py /path/*_A.png /path/*_B.png outputFolder

Note: The pair of images must have different names as they are exported in the same folder.