Improve feature extraction
MINDTCT extracts only minutiae information such as ridge end or ridge bifurcation.
Unfortunately it's not enough for accurate fingerprint matching for sensors with small area.
We can implement extraction of representative ridge points [1] or ridge shape features to increase number of feature points
[1] https://cedar.buffalo.edu/~srihari/papers/RRP-full.pdf [2] https://www.sciencedirect.com/science/article/abs/pii/S0957417417304360