The image registrator uses FAST feature finder and DAISY feature descriptor for registration.
It can align images of different size by padding them with 0 values.
The image registrator can work with multichannel grayscale TIFFs and TIFFs with multiple z-planes.
Images MUST have OME-TIFF XML in their description.
The script does linear image registration. To avoid excessive memory consumption it extracts features from tiles instead of a whole image.
-i paths to images you want to register separated by space.
-r reference image id, e.g. if -i 1.tif 2.tif 3.tif, and you ref image is 1.tif, then -r 0 (starting from 0)
-c reference channel name, e.g. DAPI. Enclose in double quotes if name consist of several words e.g. "Atto 490LS".
-o directory to output registered image.
-n multiprocessing: number of processes, default 1
--tile_size size of a side of a square tile used for registration, e.g. --tile_size 1000 = tile with dims 1000x1000px
--num_pyr_lvl number of pyramid levels. Default 3, 0 - will not use pyramids
--num_iter number of registration iterations per pyramid level. Default 3, cannot be less than 1
--stack add this flag if input is image stack instead of image list
--estimate_only add this flag if you want to get only registration parameters and do not want to process images.
--load_param specify path to csv file with registration parameters
python reg.py -i "/path/to/image1/img1.tif" "/path/to/image2/img2.tif" -o "/path/to/output/directory" -r 0 -c "Atto 490LS" -n 3
numpy pandas tifffile opencv-contrib-python scikit-learn scikit-image
scikit-image is necessary for affine transformation of big images that has more than 32000 pixels in one or two dimensions.
The affine registration process in scikit-image requires usage of float64 data, so you need amount of RAM at least 3 times the size of the picture (channel, z-plane).