A new algorithm for fingerprint feature extraction without the necessity to improve its image

Based on Gabor filter, an algorithm is worked out and presented in this work. The algorithm uses ridge endings and ridge bifurcation to represent a fingerprint image. The experimental results have proven the algorithm completion in preparing the fingerprint image for simple classification and hence high success rate of recognition. Spurious features from detected set of minutiae are deleted by a postprocessing stage. The detected features are observed to be reliable and accurate. The algorithm was implemented in Matlab and therefore it is under steady modification and improvement as each step can be easily visualized graphically to check for further analysis. The best feature of the algorithm is the unnecessity for noise removal, brightness or contrast improvement, normalization or even histogram equalization.