Improvement in Minutiae Detection by Single Ridge Local Analysis for Fingerprint Image Processing

This paper concerns algorithms related to analysis of fingerprint images and can be useful in the areas of image preprocessing and fingerprint matching based on extraction of minutiae. Algorithms are based on efficient application of detailed information about ridges in fingerprint patterns. The main goal of this algorithm is to optimize the so called ridge-following algorithm in grayscale images for detection of minutiae. Proposed modifications make use of local characteristic features from patterns of neighboring parallel ridges in order to minimize probability of skipping over ridges in a process of segmentation of a single ridge. This is a crucial problem in analysis of low quality images acquisited from a real sensor. The method shows to be efficient in proper minutiae detection together with other methods in fingerprint image processing. The solution of the use of local characteristic features proved to be useful also for filtration and segmentation of fingerprint images.