Applying Multifractal Spectrum Theory to Fingerprint Features Recognition

Fingerprint features recognition which can be used to distinguish between individuals is an intriguing study with many potential applications. In this paper, a new method for fingerprint recognition based on multifractal spectrum theory was proposed. The recognition process can be divided into the following main steps: (1) Extracting the core point in fingerprint; (2) Fragmenting the fingerprint image to get a subimage with fixed size; (3) Thinning the fingerprint image by using an improved OPTA algorithm; (4) Segmenting the curves in fingerprint image into digital straight segments with normalized straight length threshold; (5) Selecting the appropriate dividing scale to segment the processed fingerprint image; (6) Calculating and analyzing the multifractal spectrum curve ) (  f  ; (7) Fitting curve equation and extracting the characteristic parameters of ) (  f  ; (8) Finally, the parameters matching and fingerprint feature recognition. A large number of experimental results show that our method is effective.