A new fingerprint ridges frequency determination method

Abstract

Abstract:In many cases the average fingerprint ridge frequency in the whole fingerprint image should be determined. This parameter is, among other things, very important in procedures, where so-called fingerprint reference point has to be located. Unfortunately, as far, ridges frequency has been determined a priori, on the basis of anthropological measurements of the human population. This parameter is established as 10 pixels, regardless of captured fingerprint image resolution. Until now it is assumed, that this resolution is equal to 500 dpi. In practice, scanners obviously work with different resolution. This point of view has been revised and a new method of fingerprint ridge frequency measurement was introduced. This method allows to precisely calculate fingerprint ridges frequency -individually for any fingerprint image. The new approach allows determining more precisely, among other things, so-called fingerprint reference point, which is an important element of fingerprint recognition procedures.

Fingerprint Ridge Distance Estimation: A Mathematical Modeling

“…Ridge distances of three to 20 pixels, and three to 25 pixels, respectively are stated to be reasonable values to be used in typical 500 dpi fingerprint images [3,6]. Adjacent ridges are expected to be located eight to 12 pixels apart [14] in 500dpi images and 10 pixels has been observed as the average ridge distance [6,7,14,15,16]. In this paper, the range T used was T min =5 pixels and T max =20 pixels, which was based on half and doubling of the average ridge frequency 1/T which was also used in [16].…”

“…It detects all possible peaks and calculates the intervals in the histogram as the ridge distance. Other similar methods that used peak intervals as well as valley intervals by utilizing the wavelike signature of the fingerprint in the estimation of the ridge distance within an oriented local window can be found in [6,7,8]. From the work in [9], mathematical characterization using 2D sinusoidal signal has been used.…”

“…A spatial domain method. The ridge lines are either projected along its orientation direction or its grey-levels profile is found in the direction orthogonal to the orientation [1,3,5,6,7]. This method is used due to the periodic nature of the ridges.…”