Li S.Z., Jain A.K. (eds.) Encyclopedia of Biometrics

Подождите немного. Документ загружается.

according to different ﬁnger placements over the sens-

ing element (see Fig. 13 ). In fact, due to the skin

plasticity, the application of forces, some of whose

components are not orthogonal to the sensor surface,

produces non-linear distortions (compression or

stretching) in the acquired ﬁngerprints (see

▶ Finger-

print Matching).

In ‘‘Synthetic Fingerprint Generation’’ [14 ], the

skin-distortion model introduced by Cappelli, Maio,

and Maltoni [15] is exploited. While in the latter, the

distortion model was applied to re-map minutiae

points, in order to improve ﬁngerprint matching,

Fingerprint Sample Synthesis. Figure 11 Three impressions of the same real finger as captured when the finger is

dry, normal and wet, respectively.

Fingerprint Sample Synthesis. Figure 12 The application of different levels of erosion/dilation to the same ridge line

pattern.

Fingerprint Sample Synthesis. Figure 13 Two

impressions of the same real finger where a few

corresponding minutiae are marked to highlight

distortion.

520

Fingerprint Sample Synthesis

here the mapping has to be applied to the whole image,

in order to simulate realistic distorted impressions. In

Fig. 14, a ridge line pattern and its distorted impres-

sion are shown.

Noising and Global Translation/Rotation

During ﬁngerprint acquisition, several issues contrib-

ute to deter iorate the original signal, thus produc-

ing a gray-scale noisy image: irregularity of the

ridges and their different contact with the sensor sur-

face, presence of small pores within the ridges, presence

of very-small-prominence ridges, gaps and clutter-

ing noise due to non-uniform pressure of the ﬁnger

against the sensor. Furthermore, the ﬁngerprint is

usually not perfectly centered in the image

and can present a certain amount of rotation. The

noising phase sequentially performs the following steps:

1. Isolate the valley white pixels into a separate layer.

This is simply performed by copying the pixels

brighter than a ﬁxed threshold to a temporary

image

2. Add noise in the for m of small white blobs of

variable size and shape. The amount of noise

increases with the inverse of the ﬁngerprint border

distance

3. Smooth the resulting image with a 3  3 averaging

box ﬁlter

4. Superimpose the valley layer to the image obtained

5. Rotate and translate the image

Steps 1 and 4 are necessary to avoid an excessive overall

image smoothing. Fig ure 15 shows an example where

the intermediate images produced after steps 2, 4 and 5

are reported.

Background Generation

The output of the previous step is a ﬁngerprint that

appears realistic, but the image background is

completely white. In order to generate backgrounds

Fingerprint Sample Synthesis. Figure 14 A synthetic ridge line pattern (on the left) and a distorted impression (on the

right); the equivalent distortion of a square mesh is shown in the middle. To better highlight the non-linear deformations,

some corresponding minutiae are connected by white segments in both the fingerprint images.

Fingerprint Sample Synthesis. Figure 15 An example

of noising and global translation/rotation, where the

intermediate images produced after steps 2, 4 and 5 are

reported.

Fingerprint Sample Synthesis

521

similar to those of ﬁngerprints images acquired with a

given sensor, a statistical model based on the KL trans-

form (see

▶ Dimensionality Reduction) is adopted.

The model requires a set of background-only images

as a training set (see Fig. 16): a linear subspace that

represents the main variations in the training back-

grounds is calculated and then used to randomly gen-

erate new backgrounds.

Figure 16 shows some examples of the background

images (obtained from an optical acquisition sensor)

used as a training set for the background generation

step; Fig. 17 reports three synthetic ﬁngerprints with

backgrounds generated according to the above-

described model.

Related Entries

▶ Anatomy of Fingerprint

▶ Biometric Sample Synthesis

▶ Fingerprint Classiﬁcation

▶ Fingerprint Databases and Evaluation

▶ Fingerprint Features

▶ Fingerprint Singularities, Minutiae, Pores

References

1. UK Biometrics Working Group: Biometric Evaluation Method-

ology (2002)

2. Biometric Information Management and Security, American

National Standards Institute, X9.84, 2001

3. Cappelli, R., Maio, D., Maltoni, D., Wayman, J.L., Jain, A.K.:

Performance evaluation of ﬁngerprint veriﬁcation systems. IEEE

Trans. Pattern Anal. Mach. Intell. 28(1), 3–18 (2006)

4. Ku

cken, M.: Models for ﬁngerprint pattern formation. Forensic

Sci. Int. 171(2–3), 85–96 (2007)

5. Sherstinsky, A., Picard, R.W.: Restoration and enhancement of

ﬁngerprint images using M-lattice – a novel non-linear dynami-

cal system. Proceedings of the 12th International Conference on

Pattern Recognition, Jerusalem, pp. 195–200 (1994)

6. Ku

cken, M., Newell, A.C.: A model for ﬁngerprint formation.

Europhys. Lett. 68(1), 141 (2004)

Fingerprint Sample Synthesis. Figure 16 Examples of background-only images (acquired from an optical sensor)

used for training the background generator.

Fingerprint Sample Synthesis. Figure 17 Three synthetic images with backgrounds generated according to the

model in [14].

522

Fingerprint Sample Synthesis

7. Cummins, H.: Epidermal-ridge conﬁgurations in developmental

defects, with particular reference to the ontogenetic factors

which condition ridge direction. Am. J. Anat. 38, 89–151 (1926)

8. Bonnevie, K.: Studies on papillary patterns in human ﬁngers.

J. Genet. 15, 1–111 (1924)

9. Kosz, D.: New numerical methods of ﬁngerprint recognition

based on mathematical description of arrangement of dermato-

glyphics and creation of minutiae. In: Mintie, D. (ed.)

Biometrics in Human Service User Group Newsletter, (1999)

10. Bicz, W.: “The idea of description (reconstruction) of ﬁnger-

prints with mathematical algorithms and history of the develop-

ment of this idea at Optel,” (Optel, 2003). http://www.optel.pl/

article/english/idea.htm. Accessed 18 Dec 2007

11. Sherlock, B.G., Monro, D.M.: A model for interpreting ﬁnger-

print topology. Pattern Recognit. 26(7), 1047–1055 (1993)

12. Cappelli, R., Erol, A., Maio, D., Maltoni, D.: Synthetic ﬁnger-

print-image generation. Proceedings of the 15th International

Conference on Pattern Recognition, vol. 3, pp. 475–478. Barce-

lona (2000)

13. Cappelli, R., Maio, D., Maltoni, D.: “Synthetic ﬁngerprint-

database generation.” Proceedings of the 16th International

Conference on Pattern Recognition, vol.3, pp. 744–747. Que

bec

City (2002)

14. Cappelli, R.: Synthetic ﬁngerprint generation. In: Maltoni, D.,

Maio, D., Jain, A.K., Prabhakar, S. (eds.) Handbook of Finger-

print Recognition. Springer, New York (2003)

15. Cappelli, R., Maio, D., and Maltoni, D.: “Modelling plastic

distortion in ﬁngerprint images”. Proceedings of the Second

International Conference on Advances in Pattern Recognition

(ICAPR2001), Rio de Janeiro, pp. 369–376 (2001)

Fingerprint Scan

▶ Biometric Sample Acquisition

Fingerprint Sensor

▶ Fingerprint, Palmprint, Handprint and Soleprint

Sensor

Fingerprint Signatures

▶ Fingerprint Features

Fingerprint Singularity

Fingerprint singularity is deﬁned as a core or delta of a

ﬁngerprint. Minutiae have dual correspondence

between the nor mal image and density-inverted

image, that is, terminations appear as bifurcations

and vice versa. However, singularity does not have

such trait.

▶ Fingerprint Image Enhancement

Fingerprint Skeletonization

Fingerprint skeletonization, also referred to as thin-

ning, is the process of reducing the width of binarized

ridgelines to 1 pixel. Standard thinning algorithms are

applicable. Modiﬁed methods based on local ridge

orientation have been proposed to improve skeletoni-

zation accuracy. Post-processing for skeleton image,

such as skeleton adjustment, is also important.

▶ Fingerprint Image Enhancement

Fingerprint Templates

WEI-Y UN YAU

Institute for Infocomm Research, Agency for Science,

Technology & Research, Singapore

Synonym

Fingerprint representation

Definition

A ﬁngerprint template is a set of stored ﬁngerprint

features extracted from the ﬁngerprint of a user.

It is stored during the enrollment process to represent

the actual owner of the ﬁngerprint. It is subsequen-

tly compared directly to the ﬁngerprint features of

the query ﬁngerprint in order to establish whether the

Fingerprint Templates

523

query ﬁngerprint is obtained from the same person as

the actual owner. It should be noted that the original

ﬁngerprint or its enhanced or compressed form is not

a ﬁngerprint template.

Introduction

As discussed in the section on general biometrics, the

operation of a ﬁngerprint recognition system, just like

any other biometric system, follows a common process

ﬂow a s shown in Fig. 1.

A ﬁngerprint sensor is required to capture the

ﬁngerprint image which is then processed by a feature

extractor to obtain the unique features of the ﬁnger-

print. If the user is new, the features extracted are

stored in a database, typically along with other person-

al details of the new user such as name, and identiﬁca-

tion num ber. This set which stores the ﬁngerprint

feature is commonly known as a ﬁngerprint template.

The process by which this is done is called the enroll-

ment process. Subsequently, when the user wants to use

the ﬁngerprint recognition system, the ﬁngerprint fea-

tures extracted from the ﬁngerprint image acquired

live or as a query image provided into the system are

compared against the stored ﬁngerprint template(s).If

the comparison involves only one ﬁnger print template

from the database, such as when the user key in the

name to retrieve the enrolled ﬁngerprint template, the

comparison process is called veriﬁcation. Alternatively,

the comparison can be done against all the ﬁngerprint

templates stored in the database and such a process is

referred to as identiﬁcation.

Composition of Fingerprint

Template

In general, a ﬁngerprint template contains the unique

features extracted from the ﬁngerprint image.

However, the exact content varies according to the

type of algorithm used to extract and match the ﬁnger-

print. Nevertheless, if the stored ﬁle is merely the en-

hanced or compressed ﬁngerprint image or the original

ﬁngerprint image itself, it is not considered a ﬁngerprint

template. There are two general types of algorithm used

in ﬁngerprint feature extraction and matching, namely,

minutia-based and pattern-based or ridge feature-based

[1, 2]. Minutia arises when a ﬁngerprint ridge comes to

an end (called ridge ending) or when it forks out into

two ridges (called bifurcation). A sample ﬁngerprint

image with the detected minutiae is presented in Fig. 2.

Minutia detection is a complex process and is thus

beyond the purview of this contribution.

Each minutia, F, can be represented by a parameter

vector F = (x, y, ’,t)

where (x, y) is the coordinate in

the image, ’ the local ridge direction and t the type of

the minutia (i.e., bifurcation or ridge ending). The

basic composition of the minutia template, S,ofa

ﬁngerprint image is then the set of all n valid minutia

parameter vectors found in the ﬁngerprint image

given by:

S ¼ F

¼ x

; y

;’

; t



; k ¼ 1; 2; :::; n ð1Þ

Apart from the minutia information, the ridge count

[2] between two minutiae, which is the number of

ridges inte rsecting a straight line joining two minutiae,

is commonly used and included in the template. The

non-minutia data commonly extracted and included

in the template are the location, direction and the

number of core and

▶ delta points. There are many

other details that can be extracted and included in the

template such as a short ridge line information asso-

ciated to the minutia, and the number and type of

minutia encountered by the straight line used in the

ridge count with the aim to improve the performance

of ﬁngerprint matching.

A popular pattern- based approach is the Finger-

Code [3] approach. The ﬁngerprint image is tessellated

Fingerprint Templates. Figure 1 Process Flow of a Fingerprint Recognition System.

524

Fingerprint Templates

into sectors and bands with respect to a reference

point, such as the

▶ core point. A bank of Gabor ﬁlters

is then applied to each cell in the tessellation, bounded

by the boundary delineated by the sector and band of

the tessellation. For every cell, the average absolute

deviation of each ﬁlter response over all the pixels in

the cell is computed and used as an element of the

feature vector, called the FingerCode. To approximate

rotation invariant, rotate the FingerCode cyclically.

Thus, the combined FingerCode becomes the template

for this approach. Another proposed approach

includes dividing the ﬁngerprint image into small

blocks. For each block, the spectral information that

describes the ﬁngerprint pattern in that block as closely

as possible is obtained. This can be done using Discrete

Fourier Transform, Gabor Filterbanks or by selecting

the spectral component from a predeﬁned set of spec-

tral triplets [ 4] (see Fig. 3). The parameters describing

the spectral component for each block, such as (y, l, d)

for the spectral triplets, is quantized to limited discrete

values and then stored as a feature vector. The set of

feature vectors for all the blocks in the region of inter-

est of the ﬁngerprint image is then stored as the ﬁnger-

print template.

Storage of Fingerprint Template

Usually ﬁngerprint templates are stored in a central

database residing in a central database server. To per-

form a match, the extracted features from the query

ﬁngerprint image are sent to the server. Such a model

requires connection from the point where the query

image is acquired to the central server. Since the ﬁn-

gerprint template is stored in a central server, it carries

a notion of ‘‘big brother’’ which is a cause of concern

for the privacy advocates as such a system is capable of

tracking an individual. Another model for the storage

of ﬁngerprint template is the distributed database con-

cept. The ﬁngerprint templates are stored at each unit

where the use of the ﬁngerprint system is the most

common. However, if a user wishes to be recognized

in the other system, the ﬁngerprint template has to be

sent to the unit in advance, usually via a central server

which acts as a backup and synchronization unit. Such

a model is usually preferred to the central database

model if the usage spans over a large geographical

area. Instead of depending on a database, the ﬁnger-

print template can also be stored in a token such as a

smartcard, memory stick or thumb drive in a fully

decentralized model. A smartcard is usually preferred

since it is generally regarded as more secure. The user

carries the token containing his or her ﬁngerprint

Fingerprint Templates. Figure 3 A Sample Set of

Spectral Triplets Representation (right) of a Fingerprint

Image Block (left).

Fingerprint Templates. Figure 2 A Fingerprint Image Showing Detected Ridge Endings and Bifurcations.

Fingerprint Templates

525

template with him/her. To perform ﬁngerprint match-

ing, the user has to present the token and the ﬁnger-

print template stored in the token is retrieved for

matching with the query features. Alternatively, the

matching is performed inside the token itself with the

query features sent into the token. As such, the use of

the ﬁngerprint system is not dependent on any con-

nectivity. Also, the number of users can easily be scaled

when managing the template database. Unfortunately,

once the token is lost, the genuine user is unable to use

the system.

Template Synthesis

Various types of ﬁngerprint sensors are available that

capture live ﬁngerprint images. Traditionally, those

used for law enforcement purposes require a sensor

with a sensing area of at least 2.54cm  2.54cm. The

consumer version is usually smaller, about a quar ter of

the size or even smaller. However, such a small sensor is

not able to image the complete portion of the ﬁnger

that touches the sensor. Consequently, if the user does

not position his/her ﬁnger such that the contact por-

tion of the skin is largely similar to the portion used

during enrollment, then the matching will fail. This

will result in false non-match, causing inconvenience

to the user. To solve this, an image mosaicking tech-

nique has been proposed [5] for constructing a com-

posite ﬁngerprint from an image sequence of partial

ﬁngerprints. This is done by applying a low pass ﬁlter

to smooth the images and then compressing the intensi-

ty to the range of [10, 20]. The images are then aligned

using the Iterative Closest Point algorithm [6] before

superimposing the aligned images to form the

▶ mosaic.

Another mosaicking technique for rolled ﬁngerprints

has been proposed in [7]. Alternatively, a minutia-

based template synthesis approach to combine the

various ﬁngerprint templates obtained from the small

ﬁngerprint images into a composite template which

resembles the template obtai ned using a larger ﬁnger-

print image has been developed in [8].

Minutia-based template synthesis is performed by

ﬁnding all the correspondent or matched minutiae

between two ﬁngerprint images, I

and I

,tobe

synthesized. Based on the matched minutiae, an afﬁne

transformation that maps the minutiae from I

to I

then determined. This is repeated until all the other

ﬁngerprint images are synthesized. The experimental

comparison [9] revealed that the template synthesis

approach is faster, less affected by elastic deformation,

and is more suitable for larger partial images while the

image mosaicking approach is more appropriate when

accurate performance for small partial images is

required.

Template Improvement

Fingerprint images are often corrupted by noise, imag-

ing artifacts a nd affected by the skin condition (wet,

dry), amount of pressure exerted when touching the

sensor, etc. This causes the occurrence of missing min-

utiae (valid minutiae are not detected) or spurious

minutiae (false minutiae detected). Thus, accurate

detection of minutia is a very challenging task. If many

dropped or spurious minutiae are present in the ﬁnger-

print template, the usability of the ﬁngerprint recogni-

tion system is affected. To expect the user to re-enroll

regularly may cause a lot of convenience. The purpose

of template improvement is to improve the ﬁngerprint

template using multiple ﬁngerprint images captured

over a period of time [10]. For each minutia in the

template, it is initialized with a default certainty level.

When a query ﬁngerprint submitted after a time inter-

val is matched above a predeﬁned threshold, the tem-

plate and the certainty level associated with each

minutia will be updated. This is done by ﬁnding

those unmatched minutiae in the template within the

region which overlaps with the query ﬁngerprint and

then reducing their certainty level by a predeﬁned

weight, a. Next, all the unmatched minutiae found in

the query ﬁnge rprint outside of the overlapping region

will be included in the template using the template

synthesis technique but with a reduced certainty level

of (1  a). Then all minutiae in the template with a

certainty level lower than a predeﬁned threshold will

be removed. In this way, spurious minutiae can be

eliminated after many unmatched iterations while the

missing minutiae can be incorporated.

Template Interchange

There are many ways in which a ﬁngerprint system can

generate a ﬁngerprint template. In order to facilitate

the interchanging of ﬁngerprint templates among the

526

Fingerprint Templates

various ﬁngerprint systems from different vendors,

ﬁngerprint templates have to be coded in a consistent

manner. This is deﬁned by the International Organiza-

tion for Standardization (ISO) and the International

Electrotechnical Commission (IEC). The ISO/IEC

documents specify the standards for the minutia-

based templa te [11], the pattern-based template [4],

and the combined minutia and pattern template, called

the pattern skeletal [12]. The standards specifying con-

sistent formats for the construction of ﬁngerprint tem-

plates comprise 3 sections:

1. A header which describes the generic information

about the template and its source.

2. A normative section which describe s all the man-

datory features to be included and the manner in

which they have to be coded.

3. A non-normative section which allows for other

non-mandatory information to be included in the

ﬁngerprint template.

Image Reconstruction from Template

It is often assumed that the minutiae-based template

cannot be used to construct back the corresponding

original ﬁngerprint, partly because the way the data is

stored in the template is proprietary. However, it has

been shown that this is possible with a standard tem-

plate [13], such as those deﬁned by the ISO/IEC [9].

The general idea is to reconstruct the orientation pat-

tern using the orientation modeling approach [14] and

the ﬁngerprint area based on the template information.

Subsequently, the ridge pattern is developed by apply-

ing a high gain Gabor ﬁlter adjusted to the local fre-

quency and orientation and then rendering it to make

the image look realistic [2].

Summary

A ﬁngerprint template contains the unique features of

a ﬁngerprint image and can be used for ﬁngerprint

matching. The exact composition of the template is

dependent on the algorithm used to extract the unique

features. Nevertheless, international standards exist to

facilitate the interchanging of the template. It can be

stored in a database which can either be centrally

managed, distributed or stored in po rtable tokens

instead of a database. Template synthesis and template

improvement techniques can be used to improve the

performance of the system when dealing with small

ﬁngerprint sensor and poorly enrolled ﬁngerprint tem-

plates respectively.

Related Entries

▶ Enrolment

▶ Fingerprint matching

▶ Minutia

References

1. Jain, A.K., Hong, L., Pankanti, S., Bolle, R.: An Identity-authen-

tication System Using Fingerprints. Proceedings of the IEEE

85(9), 1365–1388 (1997)

2. Maltoni, D., Maio, D., Jain, A.K., Prabhakar, S.: Handbook of

Fingerprint Recognition. Springer-Verlag, New York (2003)

3. Jain, K., Prabhakar, S., Hong, L., Pankanti, S.: Filterbank-based

Fingerprint Matching. IEEE Trans. Image Process. 9(5), 846–859

(2000)

4. Standards document ISO/IEC 19794–3: Biometric Data Inter-

change Formats – Part 3: Finger Pattern Spectral Data

5. Jain, A.K., Ross, A.: Fingerprint Mosaicking, In: Proceedings of

the International Conference on Acoustic, Speech Signal Proces-

sing, vol. 4, pp. 4064–4067. Washington, DC (2002)

6. Besl, P.J., McKay, N.D.: A Method for Registration of

3D Shapes, IEEE Trans. Pattern Anal. Mach. Intell. 14(2),

239–256 (1992)

7. Bolle, R.M., Ratha, N.K., Connell, J.H.: Image Mosaicking

for Rolled Fingerprint Construction, In: Proceedings of Inter-

national Conference on Pattern Recognition, vol. 2, pp.

1651–1653 (1998)

8. Yau, W.Y., Toh, K.A., Jiang, X.D., Chen, T.P., Lu, J.W.: On

Fingerprint Template Synthesis, In: Proceedings of the 6th Inter-

national Conference on Control, Automation, Robotics and

Vision 5–8 (2000)

9. Moon, Y.S., Yeung, H.W., Chan, K.C., Chan, S.O.: Template

Synthesis and Image Mosaicking for Fingerprint Registra-

tion: An Experimental Study, In: Proceedings of the IEEE Con-

ference on Acoustics, Speech, Signal, vol. 5, pp. V-409–V-412

(2004)

10. Jiang, X., Ser, W.: Online Fingerprint Template Improve-

ment, IEEE Trans. Pattern Anal. Mach. Intell. 24(8), 1121–1126

(2002)

11. Standards document ISO/IEC 19794–2: Biometric Data Inter-

change Formats – Part 2: Finger Minutiae Data

12. Standards document ISO/IEC 19794–8: Biometric Data Inter-

change Formats – Part 8: Finger Pattern Skeletal Data

Fingerprint Templates

527

13. Cappelli, R., Lumini, A., Maio, D., Maltoni, D.: Fingerprint

Image Reconstruction from Standard Templates, IEEE Trans.

Pattern Anal. Mach. Intell. 29(9), 1489–1503 (2007)

14. Vizcaya, P., Gerhardt, L.: A Nonlinear Orientation Model for

Global Description of Fingerprints, Pattern Recognit. 29(7),

1221–1231 (1996)

Fingerprint Thinning

Obtaining a 1 pixel wide digital skeleton of ridges.

▶ Fingerprint Features

Fingerprint, Forensic Evidence of

DIDIER MEUWLY

Netherlands Forensic Institute, The Hague,

The Netherlands

Synonyms

Fingermark identiﬁcation procedure; Automatic ﬁnger-

print identiﬁcation system; Forensic evaluation of

ﬁngerprints and ﬁngermarks.

Definition

Forensic evidence of ﬁngerprint is the ﬁeld of forensic

expertise related to the inference of the identity of

source from the examination of all the friction ridge

skin, namely the ﬁngers, the palms, the toes , the soles,

and their marks. But for the sake of simplicity, the text

is mainly focused on ﬁngerprints and ﬁngermarks. The

extreme variability of the ﬁngerprints derives ﬁrstly

from the knowledge of the morphogenesis of the

papillary ridges pertaining to embryology and, second-

ly, from statistical researches per taining to dactylo-

scopy. This variability is mainly used in four different

processes within forensic science: identity veriﬁcation,

forensic intelligence, forensic investigation, and foren-

sic evaluation. The ﬁrst three processes are based on

the use of Automatic Fingerprint Identiﬁcation Sys-

tems (AFIS). The fourth process, forensic evaluation, is

an expert-based process, builton procedure, training, and

experience. The procedur e and practice vary a lot between

countries, principally regarding the thres hold used for

forensic identiﬁcation. Most of the European and South

American countries favor a quantitative approach based

on a numerical standard when the USA, UK, and most of

the Scandinavian countries have adopted a qualitative

approach based on the experience and knowledge of the

dactyloscopist. For both approaches, the decision is an

expert opinion that is deterministic: exclusion, inconclu-

sive or identiﬁcation. As the current practice is not error-

free and partly based on the subjective probabilities of

the dactyloscopists, efforts are made to develop a new

approach based on a logical inference model and statisti-

cal probabilities, in order to assist the dactyloscopists in

producing a logical, testable, and quantitative evaluation

of the ﬁngerprint evidence.

Nomenclature

At the end of the 19th century William Herschel and

Henry Faulds expressed the principles of the forensic

use of ﬁngerprints and ﬁngermarks: the use of ﬁnger-

prints and ﬁngerprint databases for the identiﬁcation of

serial offenders and the use of ﬁngermarks to establish a

link between a crime scene or an object and an individ-

ual. In literature, confusion exists between the term

ﬁngerprint and ﬁngermark. This article uses a uniform

terminology: the ﬁnger dermatoglyphics and their stan-

dard rolled inked impressions are named ﬁngerprints,

whereas recovered traces left by unprotected ﬁngers are

named ﬁngermarks. In criminal records, reference

prints are collected using forms named ten-print cards.

Individuality of the Fingerprint

Confusion surrounds the terms identity, identify, and

identiﬁcation in forensic science. This is clearly demon-

strated in popular practice, when the perpetrator of an

infringement is said to be ‘‘identiﬁed from her/his

ﬁngerprints’’. The perpetrator is not identiﬁed, but

individualized. What is proved by the ﬁngerprints is

individuality. To individualize a human being on the

basis of ﬁngermarks in forensic science ultimately con-

sists in determining if an individual is the source of

528

Fingerprint Thinning

the ﬁngermark linked to the criminal activity [1]. The

individuality of ﬁngerprints derives ﬁrstly from

the knowledge of the morphogenesis of the papillary

ridges pertaining to embryology and, secondly, from

statistical researches pertaining to dactyloscopy.

Morphogenesis

The friction ridge skin morphogenesis offers a

biological basis to explain the variability in friction

ridge patterns. The morphogenesis of the human

hands and feet star ts during the 6th week of the esti-

mated gestational age (EGA). The pattern of ridge skin

is established from the 10th week to the 14th week of

EGA when the basal layer of the volar epidermis

becomes folded and for ms the primary ridges. This

process is inﬂuenced by the volar pads, local eminences

of subcutaneous tissue in well-deﬁned locations of the

volar surfaces. It is conjectured that the inversion of

the volar pads creates tensions in the epidermis that

align the ridge pattern [2]. From this moment on up to

the 16th week of EGA, the tissues growing under the

dermis, named volar pads, induce physical stress in the

cell layers constituting this dermis. This physical stress

forms a two-dimensional structure of ridges on the

palms, the soles, the ﬁngers tips, and the toes. From

the 16th to the 24th week of EGA, the dermis matu res;

secondary dermal ridges start to develop between the

primary dermal ridges and bridges, named dermis

papillae, appear between the apex of the primary and

secondary ridges. After 24 weeks of EGA, the develop-

ment of the dermis is ﬁnalized and the epidermis is

gradually formed by cell development from the dermis,

named papillary ridges. In its ﬁnal stage, the papillary

ridges grow as a three-dimensional structure based on

the two-dimensional pattern. The anchorage of this

epidermal structu re in the dermis ensures the stability

and the permanence of the dermatoglyphics. Therefore

a permanent modiﬁcation or destruction of the der-

matoglyphics can only occur in case of destruction of

the dermis [3].

Variability of the Fingerprint

The ﬁngerprint is expressed through the interaction of

genotype, development, and environment; therefore

this biometric modality is qualiﬁed as epigenetic, sim-

ilar to the iris of the eye but contrarily to a DNA

sequence, from which by instance a DNA proﬁle is

extracted, that is genetically determined. The informa-

tion content in the ﬁngerprint ridges is structured in

three levels named the general pattern, the minutiae,

and the third level details.

General Pattern

The general shape of the ridge ﬂow, named general

pattern, is to some degree indirectly genetically inher-

ited and is classiﬁed in three generic types: arches

(simple or tented ), loops (left or right), and whorls

(including various composite forms). The approxi-

mate center of the general pattern is named the core,

and the small area where 3 ﬂows of ridges me et to form

a triangular pattern is called a delta. Arches have no

delta, loops have 1 delta, and whorls, 2 deltas (Fig. 1).

Minutiae

In addition to the general ridge ﬂow of the ridges,

deviations appear along the papillary ridges. They are

named minutiae, and can be classiﬁed in three basic

types: ridge ending, bifurcation, and dot (Fig. 2). All

other denominations, employed at the convenience of

Fingerprint, Forensic Evidence of. Figure 1 Examples of fingerprint with different general patterns: arch, loop, and whorl.

Fingerprint, Forensic Evidence of

529