Robust enhancement of fingerprint images obtained by ink method

Abstract

A method for fingerprint enhancement in the frequency domain based on directional log-Gabor filtering is proposed. Experimental results show that the method can successfully enhance even low-quality fingerprint images obtained by the ink method, outperforming standard, well-known techniques including Gabor filtering.Introduction: To utilise automatic fingerprint identification systems (AFIS) in law enforcement agencies, the first step is to digitalise an archive of fingerprints obtained by the ink method. The quality of those fingerprint images greatly depends on the fingerprint acquisition device (equipment and technique) that is used, but there are also a number of factors (postnatal marks, occupational marks, creases, etc.) that contribute to a large number of spurious minutiae detected, decreasing the overall AFIS performance. That is the reason why enhancement is usually performed before minutiae extraction [1]. When dealing with a particular fingerprint image, i.e. latent, we can apply any of many digital techniques and filters, even combining and adjusting them for that particular case. However, in the case of making a database of fingerprint templates for AFIS from an archive of cards of fingerprints usually taken by ink, the process of automatic minutiae extraction is rarely intervened with by a human expert, so an enhancement method needs to be robust and applicable to a wide range of input images. In the literature, several filters have been proposed in the spatial [2] and frequency [3,4] domains. In this Letter, a robust enhancement algorithm in the frequency domain is proposed. It is based first on amplifying the dominant frequencies by a factor of raised power spectrum (noise reduction) and then by appropriate filtering with a log-Gabor filter (directional enhancement). The results prove to be preferable compared to the standard Gaborbased filtering method.

Determining the Identity of Corpses Using Fingerprints: Results from Practice and Analysis of Process Used in the Republic of Serbia

“…The two basic factors of fingerprints that are used are individually unique and permanent. The difference is that corpses' fingerprints are taken with ink on a fingerprint card, so you should pay attention to the quality of the image of the fingerprints [15][16][17]. Person recognition by comparison of fingerprint samples with other, already known prints (picture with picture), is not recommended because it often does not give an accurate result.…”

Enhanced Rotational Feature Points Matching using Orientation Correction

“…One such method is by using Gabor filters to determine feature point orientation. The Gabor filter is usually used to classify and describe texture [5] [10] [11] as well as other applications such as fingerprint verification [12], document analysis [13] and retinal identification [14]. In [15], a few kernels were generated to check the orientation of the Gabor filter.…”

Fingerprint Enhancement Based on Wavelet and Anisotropic Filtering

“…The design of accurate, interoperable and computationally light algorithms is still an open issue. 19,26,29,35,38 A¯ngerprint is an oriented pattern of ridges and valleys on the surface of ā ngertip. In an ideal¯ngerprint image, ridges and valleys alternate and°ow in a locally constant direction, as shown in Fig.…”