生物识别防伪：人脸反欺骗技术详解

需积分: 10 94 浏览量更新于2024-07-19 收藏 6.13MB PDF 举报

"Face Anti-Spoofing：数字世界的门禁守护者在信息技术的快速发展中，面部识别作为生物特征认证的一种，已经成为现代安全系统中的重要组成部分，尤其在访问控制（Access Control）逻辑层面发挥着关键作用。Face Anti-Spoofing，即面部反欺骗技术，旨在保护个人信息系统的安全性，防止通过伪造或欺诈手段获取未经授权的访问权限。逻辑访问控制（Logical Access Control）是指一系列策略和方法，用于保护个人计算机（PC）、网络和移动设备上的信息。这种控制方式广泛采用各种凭证类型，包括密码、智能卡、令牌或生物识别（如指纹、面部识别等），来验证用户身份。这些凭证分别代表用户所知（知识型），所拥有（物品型），或身体特性（生物特征型）的信息。在逻辑访问控制系统中，用户首先需进行注册，将凭证与用户身份关联起来。当用户试图访问特定资源，如应用、网络、设备或操作系统时，系统会要求用户提供一到多个验证方式。同时，系统会记录所有的访问尝试，以便于审计谁在何时对特定资源进行了访问，从而提升系统的透明度和安全性。随着密码的传统地位逐渐被多因素认证（如面部识别+密码）取代，面部反欺骗技术的重要性日益凸显。它能有效检测和阻止诸如面具、照片、视频或3D模型等欺诈手段，确保只有真正的人脸才能通过验证。例如，DigitalPersona Inc.,这家位于美国加利福尼亚州红木城的公司，就专注于提供先进的面部生物识别解决方案，强化了现代网络安全的防护屏障。总结来说，Face Anti-Spoofing是逻辑访问控制体系中不可或缺的一环，它通过先进的生物识别技术和算法，提升了用户身份验证的准确性和安全性，有效地防止了身份冒用和欺诈行为，为数字化世界的门禁管理提供了强有力的支持。随着科技的进步和用户对隐私保护意识的提高，这项技术将在未来的安全领域扮演越来越重要的角色。"

A 16 Anatomy of Friction Ridge Skin

Anatomy of Friction Ridge Skin, Fig. 2 Friction skin ridge ﬂow: whorl and loop ﬁnger patterns and unpatterned skin

from a lower ﬁnger joint. Note the minor creases (white lines), especially in the middle and right images

of minor ﬂexion creases are particularly notable

on the palm. Flexion creases form along areas

in which the skin is more strongly attached to

the underlying fascia [3]. The smallest of the

ﬂexion creases are known as white lines, which

occur randomly over the skin [1]. The prevalence

and depth of white lines increases with age.

White lines are especially prevalent on the lower

joints of the ﬁngers and on the thenar (base of

the thumb). In some cases, a large number of

white lines make the underlying ridges difﬁcult

to discern, as shown in Fig. 2.

Friction Skin Structure

Skin is a protective barrier that contains nerve

receptors for a variety of sensations, regulates

temperature, allows the passage of sweat and

sebaceous oils, and houses the hair and nails.

Friction ridge skin is differentiated from thin skin

not just by the presence of raised papillary ridges

but also by epidermis that is much thicker and

structurally more complex, by increased sensory

abilities, by the absence of hair, and by the

absence of sebaceous glands.

Skin throughout the body is composed of three

basic layers: the hypodermis, dermis, and epider-

mis.

The innermost hypodermis (also known

as subcutaneous tissue or the superﬁcial

fascia) is made of connective tissue that stores

fat, providing insulation and padding. The

hypodermis varies in thickness, but is particularly

thick in friction ridge skin [3, 4].

The dermis is composed of dense connective

tissue that provides strength and plasticity. The

dermis houses blood vessels, nerve ﬁbers and

endings, and sweat glands. In non-friction ridge

skin, the dermis also contains sebaceous (oil)

glands, hair follicles, and arrector pili muscles,

which raise hair and cause “goose bumps” [4].

The boundary between the dermis and epider-

mis is of particular interest for friction ridge skin.

The dermis and epidermis are joined by papillae,

which are columnar protrusions from the der-

mis into the epidermis, and rete ridges, which

are the areas of the epidermis surrounding the

papillae. The papillae anchor the epidermis and

contain nerve endings and blood vessels. Papillae

in thin skin are small and relatively infrequent

and are not arranged in any particular pattern.

In friction ridge skin, the papillae are densely

arranged in curved parallel lines by pairs, with

pairs of papillae surrounding the sweat pores [3]

(Fig. 3).

The epidermis provides the outermost

protective layer and is structurally very different

between friction ridge and thin skin. The epider-

mis does not contain blood vessels, and therefore,

the basement membrane, which joins the dermis

and epidermis, serves as the medium through

which nutrient and waste pass. The lowest level

of the epidermis (stratum basale) contains a

Anatomy of Friction Ridge Skin 17 A

Ridge

ending

Enclosure

Bifurcation

Dot

Sweat pores

Germinative

cells

Sweat

gland

duct

Sweat

gland

Papillae

Ridge ending

Incipient ridge

Pores Dot

Keratinized

cells

Epidermis

Dermis

Bifucation

Anatomy of Friction Ridge Skin, Fig. 3 (a) Structure of friction ridge [5]. (b) Examples of friction ridge features

single layer of basal generative cells, which are

anchored to the basement membrane. These basal

cells continuously create columns of new cells.

It takes 12–14 days for a cell to progress from

the innermost basal layer to the outermost horny

or corniﬁed layer of the epidermis. During this

time, the cell ﬂattens, forming interconnections

with the neighboring cells, is ﬁlled with keratin

(the tough protein found in hair and nails), and

dies. The dead cells are continuously exfoliated,

with the entire epidermis being renewed

approximately every 27 days. The thickest

portion of the corniﬁed layer of cells is generated

along the lines of paired papillae, resulting in

visible friction ridges, punctuated with pore

openings. The epidermis in friction ridge skin

is 0.8–1.44 mm thick, as compared to 0.07–

1.12 mm thickness elsewhere. Heavy use can

result in substantially increased thickness of the

epidermis, in the form of calluses or corns [3, 4].

Friction Skin Development

The individual characteristics of friction ridge

skin are determined during fetal development,

based on a combination of genetic and random

factors. The overall pattern of friction ridges is

determined by the formation and regression of

volar pads in the fetus. Starting at approximately

6 or 7 weeks of gestational age, human fetuses

form swellings of tissue in what will later become

the dermis: 11 of these volar pads generally

develop on each hand, with 1 at each ﬁngertip,

4 interdigital pads at the bases of the ﬁngers,

1 thenar pad at the ball of the thumb, and 1

hypothenar pad along the outside of the palm.

Each foot has 11 pads in corresponding locations.

The size, shape, and period of development of

the volar pads are determined to a large extent

by genetics. The pads continue to grow for a

few weeks and then regress as the rest of the

A 18 Anatomy of Friction Ridge Skin

hands and feet grow. The volar pads are usually

no longer evident by about 16 weeks of gesta-

tional age. During the period of volar growth

and regression, starting at about 10 weeks of

gestational age, the basal epidermal cells begin a

stage of rapid proliferation, especially surround-

ing the sweat glands. Since this process occurs

while the volar pads are regressing, the result

is that the growing cells fuse together along the

lines of stress created by the collapse of the

volar pads. While the overall form of the ridges

follows the contours of volar pads, the speciﬁc

paths, bifurcations, and endings of the ridges are

determined by the stresses encountered during

growth [1, 6, 7].

The overall form of the ridges is determined

by the topography of the volar pads, with the

pattern class determined by the height and sym-

metry of the volar pads. This can be seen most

easily in examining the areas without volar pads,

such as the lower joints of the ﬁngers and the

lower palm: the ridge ﬂow in these areas is

generally simple, with ridges ﬂowing across the

area without dramatic changes in direction. If

volar pads are small, the resulting pattern will

be an arch, with simple ridge ﬂow similar to the

areas without volar pads. If the volar pads are

large and centered, the resulting pattern will be

a whorl, with ridge ﬂow following the circuit of

the pad.

Because of the genetic basis for volar

pad formation, overall ridge ﬂow or pattern

classiﬁcation is often similar between siblings,

especially identical twins. For the same reason,

ﬁngerprint patterns on an individual’s left and

right hands are often similar to mirror images

of each other. However, because the path of any

individual ridge results from chaotic stresses,

the details of minutiae are speciﬁc to the

individual.

Prevalence

Friction skin covers the palms and soles of all an-

thropoid primates (monkeys, apes, and humans),

as well as on portions of the prehensile tails

of some new-world monkeys. Some but not all

prosimian primates (lemurs) have friction skin

on portions of their palms and soles [8]. Friction

skin is unusual in other mammals, but is found

on portions of the grasping hands and feet of

two species of tree-climbing marsupials (koalas

and one form of phalanger) [9]. Note in all cases

that friction ridge skin is associated with grasping

surfaces: the ridges increase friction, and the

greater density of nerve endings improves tactile

sensitivity.

Problems in Capturing Friction Skin

Features

Friction ridge skin is a ﬂexible, three-dimensional

surface that will leave different impressions

depending on factors including downward or

lateral pressure, twisting, and the medium used.

Even when only considering clear impressions,

the details of ﬁngerprints and palmprints vary

subtly or substantially between impressions.

As downward pressure increases, the apparent

diameter of the valleys decreases and the ridges

widen. The frequency of ridges is affected by

lateral compression or stretching. A bifurcation

of a physical ridge does not always appear as a

bifurcation in the corresponding print, but may

appear to be a ridge ending under light pressure.

Incipient ridges may become more discontinuous

or vanish altogether under light pressure. Pores

are not always evident in ﬁngerprints even

at high resolution, which can be explained

in part by the tendency to ﬁll with liquid

such as sweat or ink. This variability between

different impressions of an area of friction

skin is responsible for much of the complexity

of matching ﬁngerprints, whether performed

by human experts or automated recognition

systems.

Related Entries

 Anatomy of the Hand

 Fingerprint Classiﬁcation

 Fingerprint Recognition, Overview

 Palmprint Matching

Anatomy of the Hand 19 A

References

1. D.R. Ashbaugh, Quantitative-Qualitative Friction

Ridge Analysis: An Introduction to Basic and Advance

Ridgeology (CRC Press, Boca Raton, 1999)

2. Scientiﬁc Working Group on Friction Ridge Anal-

ysis, Study and Technology (SWGFAST): Glossary,

Version 1.0 (2003), http://www.swgfast.org/Glossary_

Consolidated_ver_1.pdf

3. S. Standring (ed.), Gray’s Anatomy: The Anatomical

Basis of Clinical Practice, 39th edn. (Elsevier, London,

2004)

4. L. Weiss (ed.), Cell and Tissue Biology: A Textbook

of Histology, 6th edn. (Urban & Schwarzenberg, Balti-

more, 1988)

5. Federal Bureau of Investigation: The Science of Fin-

gerprints, Rev 12-84. U.S. Government Printing Of-

ﬁce, Washington, DC, 1984

6. A. Maceo, Biological basis of uniqueness, persistence,

and pattern formation, in 4th International Fingerprint

Symposium, Lyon, 17–19 May 2006. http://www.

interpol.int/Public/Forensic/ﬁngerprints/Conference/

May2006/presentations/2AliceMaceoPart1.pdf,

2AliceMaceoPart2.pdf

7. K. Wertheim, A. Maceo, The critical stage of friction

ridge and pattern formation. J. Forensic Identif. 52(1),

35–85 (2002)

8. F. Ankel-Simons, Primate Anatomy: An Introduction

(Academic, San Diego, 2000)

9. M. Henneberg, K.M. Lambert, C.M. Leigh, Finger-

print Homoplasy: Koalas and Humans (Heron Pub-

lishing, Victoria, 1997), naturalSCIENCE.com

Anatomy of the Hand

Amioy Kumar

, Tanvir Singh Mundra

,and

Ajay Kumar

Biometrics Research Laboratory, Department of

Electrical Engineering, Indian Institute of

Technology Delhi, New Delhi, India

The Hong Kong Polytechnic University,

Hung Hom, Kowloon, Hong Kong

Synonyms

Hand Geometry; Hand structure

Deﬁnition

The anatomy of the human hand is quite unique

and includes the conﬁguration of bones, joints,

veins, and muscles. The physiological intercon-

nection and structure of these parts are responsi-

ble for the structure of the human hand. The func-

tional area of the hand includes the ﬁve ﬁngers,

palm, and wrist. Among a number of biometric

modalities that are used for human identiﬁcation,

hand-based modalities achieve high performance

and have very high user acceptance. A hand-

based biometric system integrates several physio-

logical and/or behavioral features that have their

individuality in the anatomy of the hand. The

prime focus of this study is on internal and

physiological structure of the human hand which

deﬁnes the uniqueness of various hand-related

biometric modalities.

Introduction

The anatomical study of the human hand is not

new; it dates back to prehistoric times, but it is

ﬁnding new applications in the ﬁeld of biomet-

rics. The proper understanding of the structure

requires the knowledge of function in the liv-

ing organism. As one of the basic life sciences,

anatomy is closely related to medicine and to

other branches of biology. The hands of the hu-

man being are the two multi-ﬁngered body parts

located at the end of each arm. It consists of a

broad palm with ﬁve ﬁngers, each attached to

the joint called the wrist. The back of the hand

is formally called the dorsum of the hand. The

uniqueness of the human hand, as compared to

the other animals, comes from the fact that all

the ﬁngers are independent of each other and the

thumb can make contact with each ﬁnger.

The anatomy of the hand is the key to ascertain

the individuality of hand-based biometrics. The

hand-geometry biometric largely represents the

anatomy of hand bones and muscles. The hand-

vein biometric represents the uniqueness in the

anatomy of hand veins, while the palmprint rep-

resents the epidermis on the palm. The behavioral

biometrics like signature is also highly dependent

on the anatomy of bones and muscles. There-

fore the study of hand anatomy is fundamen-

tal to ascertain the individuality of hand-based

biometrics.

A 20 Anatomy of the Hand

Structure of the Human Hand

The internal structure of the hand is an assortment

of bones, muscles, nerves, and veins.

Bones

The structure of the hand is primarily attributed

to the bones comprising the human hand. The

hand is composed of 27 bones, broadly divided

into three groups called carpals, metacarpals, and

phalanges (Fig. 1). The wrist of the hand consists

of a cluster of bones named as carpals. These

bones are considered as part of the wrist and are

responsible for the to-fro and back-forth move-

ment of the wrist. These are eight in number and

are named as:

1. Scaphoid 2. Lunate

3. Triquetrum 4. Pisiform

5. Trapezium 6. Trapezoid

7. Capitate 8. Hamate

Metacarpals are the intermediate part of the

ﬁngers and the wrist [1]. This cluster of bones

makes the central part of the hand called the

palm. The metacarpals are ﬁve in number and are

named as:

9. First metacarpal

(thumb)

10. Second metacarpal

(index ﬁnger)

11. Third metacarpal

(middle ﬁnger)

12. Fourth metacarpal

(ring ﬁnger)

13. Fifth metacarpal

(little ﬁnger)

The remaining 14 bones are called the pha-

langes. These are named as follows:

14. Proximal 15. Medial

16. Distal

There are two in the thumb and three in each

of the four ﬁngers, as shown in Fig. 2. The distal

phalanges carry the nails, the middle phalanges

are in the middle, and the proximal phalanges are

Anatomy of the Hand , Fig. 1 Skeletal structure of the

human hand

Anatomy of the Hand, Fig. 2 Phalanx bones of the

human hand

closest to the palm. The bones are the most im-

portant structure of the human hand and respon-

sible for almost all the activities of the hand. Even

so, bone structures in the hand are not a popular

candidate for biometric authentication. Being a

hidden structure of the hand, the acquisition of

hand bone images is very difﬁcult. However,

the hand bone structures are useful in forensic

identiﬁcation especially in situations when other

physiological structures are not available, e.g.,

during accidents/ﬁre. The individuality of hand

bone structures is generally believed to be low

due to the high similarity in the bone types.

剩余137页未读，继续阅读

x65255062

粉丝: 3
资源: 29

生物识别防伪：人脸反欺骗技术详解

CASIA数据库

基于tensorflow的微笑检测

Fusing Multiple Deep Features for Face Anti-spoofing.pdf单篇论文

Deep Learning for Face Anti-Spoofing A Survey.zip

HyperFAS:静默活体检测 Silent Face Anti-Spoofing Attack Detection

Learning Deep Models for Face Anti-Spoofing Binary or Auxiliary Supervision.zip

Silent-Face-Anti-Spoofing-APK

静默活体检测静脸防欺骗攻击检测silent face anti-spoofing attack detection.zip

Face-Anti-spoofing

STDN-PyTorch:这是 PyTorch 中“On Disentangling Spoof Traces for Generic Face Anti-Spoofing”的非官方实现

最新资源