Uniformity Level Approach to Fingerprint Ridge Frequency Estimation

Stages of fingerprint image enhancement include segmentation, normalization, filtering, binarization and filtering. Each of these stages has proved to be very essential for achieving a well enhanced fingerprint image. The major prerequisites to filtering a fingerprint image are ridge orientation and frequency estimations. While ridge orientation estimation is done to obtain the orientation of the ridges, ridge frequency estimation is done with a view to ascertaining the number of ridges within a unit length. The number is useful for fingerprint image filtering. In this paper, a modified fingerprint ridge frequency estimation algorithm is implemented. The modified algorithm consists of stages for estimating ridge orientation and uniformity levels. Two types of images; namely synthetic and real fingerprints were used to evaluate the performance of the algorithm. The results of the evaluation reveal that the modified algorithm shows greater speed and effectiveness than its original version. Facts also emerged on the basic characteristics of the estimates.