An image registration based on feature points Krawtchouk moments is proposed. Moments are the shape descriptors based on region. Krawtchouk moments are a set of discrete orthogonal moments and are more suitable for describing two-dimensional images compared to Zemike, Legendre moments. In the image registration based on feature points Krawtchouk moments, Krawtchouk moment invariants of the feature points neighborhood that have been extracted are solved, and then these Krawtchouk moment invariants constitute feature vectors used to describe the feature points, finally feature points are matched by calculating the Euclidean distance of feature vectors. The results of experiments show that Krawtchouk moment is simple and effective to describe image and is independent of rotation, scaling, and translation of the image.
GLAUCOMA IMAGE CLASSIFICATION USING DISCRETE ORTHOGONAL STOCKWELL TRANSFORM