ImageRegistration.m is the script for registering tif images to dicom images.
RegAlgorithm.m is the registration function.
transform_image.py is the script for applying transformation matrixes.
Images are enhanced in the first step, which includes normalization, histogram equilibrium and anisotropic diffusion. Then images features are extracted using blob detection (SURF). At last, match corresponding features in 2 images and compute the transformation matrix.
Run ImageRegistration.m and transform_image.py at the folder of dataset (keep the directory structure in USB disk).
Modules needed for registration: Matlab Computer Vision System Toolbox
Modules needed for applying transformation: opencv, pydicom