Geometry of region technique analysis for core point detection

Core point is one of the features in fingerprint image. To detect the core point location of the image, the Geometry of Region Technique is implemented in this research. There are four main steps in the technique, i.e., image normalization, estimate local orientation, calculate smoothed local orientation, and finally sine component is computed in order to obtain the core point location. Therefore, the Geometry of Region Techniques is based on the local orientation of the fingerprint image. The experiment is conducted in four scenarios based on the size of the block of an image. The experiment shows that the smaller size of the block achieves higher accuracy, since smaller size of the block provides more detail information about the orientation of the block. The highest accuracy for this method is 83%.