mm IIHIHIHIHIJIIHIIJJUHINllzlllllllUllHINHIIIHHIWII

[ 20741694
r LIBRARY “‘7’
“ Michigan State

University

 

 

 

This is to certify that the

thesis entitled

CGPARISON 0F FACIAL HDTOGRAPHS
VIA WLOGICAL ASSESSED" FOR FORENSIC SCIENCE PIRI’OSES

presented by

LAURIE AMI KRIPA

has been accepted towards fulfillment
of the requirements for

gm M

Majorfi fessor

Date W

0.7639 MS U is an Affirmative Action/Equal Opportunity Institution

g H

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINE return on or before date due.
MAY BE RECAILED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

COMPARISON OF FACIAL PHOTOGRAPHS VIA MORPHOLOGICAL

ASSESSMENT FOR FORENSIC SCIENCE PURPOSES
BY

LAURIE ANN KRUPA

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

Criminal Justice

1999

ABSTRACT

COMPARISON OF FACIAL PHOTOGRAPHS VIA MORPHOLOGICAL

ASSESSMENT FOR FORENSIC SCIENCE PURPOSES

BY

LAURIE ANN KRUPA

A facial feature classification protocol was
established for comparing facial photographs to
surveillance video stills. Twelve examiners used the

protocol, designated the morphological assessment sheet, to
narrow a population of eight photographs to determine the
value of the protocol as an exclusionary tool. The
examiners also filled out a nmmphological assessment sheet
for one specific photograph to evaluate examiner agreement
in feature descriptions. The results of the examination
indicate that morphological assessment can be a valuable
method for narrowing populations of photographs. This
benefits the law enforcement community by providing a time

saving, systematic approach to photograph comparison.

DEDICATION

I would like to dedicate this to my parents, Theodore
and Sheryl, and to my siblings, Amy and James. for their

constant support and motivation.

I would also like to dedicate this to my Elaine

Chilimigras and Marti Johnson, without whom :my time at

Michigan State would not have been the same.

ACKNOWLEDGEMENTS

I would like to thank Dr. Jay Siegel, not only
for his guidance in my graduate studies, but also for his

counsel throughout my undergraduate years.

I would also like to thank Dr. Norman Sauer, for
introducing me to the field of forensic anthropology, and

for his wisdom and patience throughout this project.

TABLE OF CONTENTS

vi.

 

 

 

 

 

 

 

 

 

LIST OF TABLES

LIST OF FIGURES vii.

INTRODUCTION

LITERATURE REVIEW

MATERIALS AND METHODS 12.

RESULTS 14.

DISCUSSION 16.

CONCLUSIONS AND RECOMMENDATIONS 18.

APPENDICES 22.

Appendix A — Tables 23.
Table 1. 24.
Table 2. 24.
Table 3. 25.
Table 4. 27.
Table 5. 29.
Table 6. 29.
Appendix B - Figures 30.

Figure l. 31.
Figure 2. 33.
Figure 3. 35.
Figure 4. 41.

REFERENCES 48.

 

 

 

 

 

 

 

 

 

 

 

 

 

Table

Table

Table

Table

LIST OF TABLES

Subject identification based on
morphological assessment sheets.

Subject identification based on
morphological assessment sheets.

Subject evaluation of photograph 1a.

Subject evaluation of photograph 1b.

vi

frontal

profile

view

view

LIST OF FIGURES

Figure l. Morphological assessment sheet for frontal
analysis.

Figure 2. Morphological assessment sheet for profile
analysis.

Figure 3. Appendix to morphological assessment sheet for
frontal analysis.

Figure 4. Appendix to morphological assessment sheet for
profile analysis.

vii

Introduction

As technology advances, surveillance video cameras
become more prevalent in commercial establishments. The
increase in the presence of surveillance cameras lends
itself to an increase in the documentation of crime (Vanezis
et al, 1996). With the rise in video evidence, there exists
a need within the law enforcement community for methods of
comparison of video still frames with photographs of
suspects.

Comparison. of video footage still frames to facial
photographs would appear to be inherently simple; however,
it is regarded as one of the most complex modes of forensic
analysis (Iscan, 1993). There are several factors which may
hamper such identifications. The significant factor is the
quality of photograph obtained from the original video
footage, which. is typically' grainy’ and. of 10W' quality.
These blurry, indistinct images can lead to difficulty in
determining feature boundaries. Moreover, the person
captured on video may be sufficiently obscured (e.g., by a
hat or shadow) to further complicate the comparison process.
Additionally, there may be alterations in the appearance of
the person in question. These changes may arise from time
lapse, whether by conscious design (e.g., altering hairstyle
or' plastic surgery) or' by' the physiological process of
agingu Differences in facial expression can result in

dramatically different photographs of the same person.

There are three primary modes of analysis currently
employed in the comparison of photographs: anthropometry,
photo-photo video superimposition, and morphological
assessment. These methods employ anthropological techniques
developed for both antemortem and postmortem identification
of individuals, as well as for medical purposes,
particularly reconstructive surgery.

While very similar to visual identification of an
actual face, photographic comparison is unique in that the
photograph of unknown origin is being compared to a
photograph of a person of known origin, rather than to an
actual individual or to human remains. This poses a unique
set of difficulties because the orientation of both items of
comparison is static and cannot be manipulated. Therefore,
particularly in video superimposition and anthropometric
analysis, it is vital that both photographs be in the same
orientation.

In order to ensure themself that a photograph is in
identical, or nearly identical, orientation to the video
still frame to which it is being compared, the analyst must
often collect and assess a large volume of photographs of
the suspect. In addition, the analyst must be able to draw
inferences from the results of the anthropometric or
superimposition analysis in order to assess the significance
of minute differences in orientation. Therefore, while both
of these tools are effective methods, the amount of time and

level of skill required pose a considerable limit on their

practicality, principally in instances involving large
numbers of suspects. The third mode of analysis,
morphological assessment, qualitatively compares the
features of one individual to another, based on the premise

that the sum of feature descriptions of an individual is

unique to that person . There are several advantages to
morphological assessment: the effect of differing
orientation or facial expression is diminished, as

morphological assessment is based on inherent assessment
abilities. With the use of a systematic protocol,
morphological assessment has the potential to be a time
saving, straightforward, and effective analytical tool for
exclusionary purposes.

The objective of this project was to develop a standard
protocol (morphological assessment sheet) for comparing
photographs as a law enforcement tool, and determine if it
is an effective instrument for narrowing a population of
suspects. The protocol was developed using anthropological
data but designed for use by law enforcement personnel; for
that reason, developmental objectives included attempting to
minimize the use of esoteric terms when possible, and
creating a supplemental instructional guide. Specific goals
in developing the protocol were to create a logical approach
for describing the face and. to minimize the amount of
subjectivity in the descriptive terms.

The method for testing the morphological assessment

sheet was to select eight photographs of Caucasian males age

18-25 (designated the ‘suspects'), and choose one photograph
as the ‘target'. A morphological assessment sheet was
filled out describing the characteristics of the target.
Copies of the filled out sheet were distributed to a panel
of twelve examiners, who were chosen to determine which
photograph the sheet was describing. The result of this
examination was that the suspect pool was narrowed in all
cases to no more than three photographs. In most cases, the
pool was narrowed to one or two photos.

These preliminary findings indicate that morphological
assessment may be an effective tool for excluding
individuals from a population of suspect photographs.

It is important to note that this is not an
identification technique, as there are numerous factors
which may potentially lead to a false positive result. The
effect of false positives in forensic science is critical,
as decisions of freedom are made based on the analyst's
findings. Therefore, when suspects are not excluded as a
result of morphological assessment, corroborative analysis
is imperative. But as an exclusionary technique,
morphological assessment will greatly benefit the law
enforcement population by serving as a straightforward
method for eliminating suspects, and thus decreasing

analysis time.

Literature Review

The three modes on analysis, anthropometry, video
superimposition, and morphological assessment have been
developed over many years for various applications.

Anthropometry, the quantitative evaluation of the human
body, involves measurement of the size, weight, and
proportions of human features (Farkas et al, 1994).
Anthropometry was first introduced as a technique for
personal identification by Alphonse Bertillon, in 1883. The
Bertillon system of identification was based on two
premises; that skeletal dimensions of humans remained fixed
from the age of twenty until death, and that skeletal size
was unique to each individual (Saferstein, 1998).
Anthropometry was considered to be the most accurate
identification technique until it was superseded by
fingerprints in the early twentieth century. Current
anthropometric research has primarily been conducted by the
medical community in the field of reconstructive surgery
(Farkas et al, 1994); however, the technique has been
employed for forensic science purposes (facial
reconstruction and photograph comparisons) as well.

In anthropometric comparison, proportionate
measurements of selected features are compared between the
known and unknown subjects. A match may be assumed if
several measurements are the same. The limitations which

are encountered in the controlled setting of the medical

community (difficulty in locating features and their
boundaries (Farkas et a1, 1994)) are exacerbated by the poor
quality of surveillance video. In their review of this
technique, Iscan et a1 (1993) cite the lack of research
specific to forensic science as its chief limitation. It is
vital that the examiner be well-versed and experienced in
this method.

Superimposition techniques were first employed to
authenticate skulls of historical figures. Initially,
comparisons were made with portraits, busts, and death masks
(Austin-Smith and Maples, 1994). The invention of
photography improved the technique of superimposition, as
antemortem. photographs eliminated. artistic interpretation
and provided a more accurate depiction of the decedent. The
first documented use of photographic superimposition for
forensic science purposes was the Ruxton murder case, in
1935, in which photographs of two women were enlarged to
life size and compared with skull tracings (Stewart, 1979).

The use of video equipment has improved the
superimposition process, both in accuracy and simplicity.
In photo-photo video superimposition, two photographs are
taken in the same orientation and each is placed beneath a
video camera. Images of each photograph are sent to an
electronic mixer and displayed on a monitor. The images can
then be superimposed over each other by fading or wiping to
determine if the features of one photo are similar to the

other. Symmetry of features, particularly features of

unique shape may be indicative of positive identification
(i.e., the subject in the first photo is the same subject in
the second) (Iscan, 1993).

Advances in technology are allowing for improvement in
the accuracy of anthropometry and superimposition. The use
of computers in taking measurements is affording greater
accuracy to the technique. One limitation noted in
anthropometry, that the photographs must be taken in
standard size, is addressed by using proportionate rather
than absolute measurements (Vanezis and Brierley, 1996).

There is a large body of research ’ devoted to
superimposition techniques (Austin-Smith and Maples, 1994;
Chai et a1, 1989; Loh and Chao, 1989; McKenna et a1. 1984,
Yoshino and Seta, 1989). The research. devoted to these
techniques is considerably applicable to facial photograph
comparisons, as it addresses such issues as determination of
focal lens length, angulation (McKenna 1984; Sekharan,1973),
and facial landmarks (Chai et a1, 1989).

Photographic superimposition has been regarded as the
least accurate of the three methods of photograph comparison
discussed by Iscan et a1 (1993). The authors cite
difficulty in obtaining photographs in identical angulation
as the primary limitation of the method, and caution against
relying on symmetry of one or two features despite a lack of
alignment of other features. Moreover, the analyst must
avoid the tendency’ of mentally' blending the images and

inferring a match where one may not exist (Venezis and

Brierley, 1996). Like anthropometry, this technique
requires a high skill level, and is time comsuming.

Morphological assessment, the technique of
qualitatively noting similarities and differences between
two objects of comparison, has a long history in forensic
science, and is used in comparative examinations of a wide
array of evidence, including fingerprints and trace
evidence. The application of this type of analysis to
facial features has been developed recently, based on
anthropological techniques.

J. Lawrence Angel introduced a list of feature
classifications (unpublished), which was later modified by
Iscan (1993), as a collection of the entire range of
morphological classifications. The intent of Iscan's work
was to "make individual variation and population differences
emerge from a seemingly unremarkable visage" and to aid the
forensic anthropologist. in. effectively' comparing' two
photographs.

Venezis et al (1996) compiled a subset of these
features to evaluate the possibility of establishing a
practical classification of the face specifically designed
for forensic purposes. This subset selected features based
on the following criteria: ease of discrimination, agreement
among’ assessors, feature jpermanence, non-reliance on
anthropometry, and that the feature be part of normal

morphological anatomical ‘variation (i.e., not injury

induced, etc.).

This classification was assessed by seven examiners,
using agreement of five or more and low feature occurrence
as an indication that the feature classification was
successful for discrimination. The results of this study,
which displayed a high level of agreement, indicated success
of the method.

These developments suggest the future potential of
anthropometry or video superimposition for identifying
purposes. The intent of this project is not to suggest that
morphological assessment is superior to alternative
techniques; analysis protocol should be determined based on
the nature of the photographs. Nonetheless, advances in
anthropometry or video superimposition will not eliminate
the time requirement for selecting suitable photographs for
comparison and for analysis. Based on the time requirement,
application of these techniques to cases with large
populations of suspects is highly impractical.
Morphological assessment is currently the most ideal
analytical mode for narrowing populations of suspects.

The results of the Venezis et al study indicate that
there is reliability in the use of feature descriptions for
exclusionary analysis. The next logical approach is to
determine whether an examiner who is presented with
surveillance video evidence can effectively use a
morphological assessment protocol to eliminate suspects from
a population” This project addresses that question. by

asking a panel of examiners to use the protocol developed

specifically for the project to attempt to eliminate
photographs from a pool of suspects which do not match the
morphological description.

As this method is designed for law enforcement
purposes, modifications must be made for use by personnel
with limited anthropological expertise. The morphological
assessment sheets proposed here subdivide features into
frontal or profile views, define or eliminate the use of
esoteric terms, provide an instructional guide with
schematic drawings for clarity, and reduce the number of
descriptive terms for features to the minimum necessary.
Moreover, the sheets address specifically the problem of
comparing photographs to surveillance video still frames,
and exclude those that would not be apparent on video (e.g.,
texture and color of hair), as well as features that would
not resolve clearly enough for evaluation (e.g., parts of
the eye). The morphological assessment protocol proposed
here includes a section for providing additional
information, such as scars, tattoos, etc. Whereas the
method of Vanezis and Brierley disregards "features regarded
as acquired anomalies resulting in many cases as a result of
injury", this method takes into account the insight from
skeletal analysis that the most reliable sources in
individuality are those features which show diversity from

pathological development, trauma, and surgical alteration.

10

However, if the morphological classification is
modified, it becomes necessary to reevaluate the
subjectivity of the feature descriptions.

This project addresses this issue in the second part of
the analysis, by having the panel of examiners each fill out
one assessment sheet for' a designated. photograph” The
sheets will be tabulated to determine the level of agreement

among the examiners for each feature.

11

Materials and Methods

The initial part of the experiment was to develop
morphological assessment sheets for frontal and profile view
analysis(Figures 1 and 2), modified from the work of Vanezis
et al (1996), and Iscan et a1 (1993). Modifications include
subdividing the facial characteristics into two sheets: one
for frontal view analysis, and one for profile. Another
limitation that was addressed. was the static nature of
Vanezis' list, by adding a section (‘other') and a facial
outline for noting the presence and location of anomalous
features. Characteristics were eliminated that were not
likely to resolve in surveillance video. Additionally, an
instructional appendix was added for each view (Figures 3
and 4), with guidelines for each feature class and a
glossary. The sheets contain features addressing the facial
shape, hairline, brow ridges, eyebrows, nose, mouth, and
chin.

Twenty professional quality photographs of Caucasian
males age 18—25 were obtained, and from these, eight were
selected based on similarity of appearance as the
‘suspects' . From the suspects population, one photograph
was randomly selected and designated the ‘target'
photograph. A morphological assessment sheet was filled out,
describing the target photograph. The photographs were then

compiled into a 'lineup' to be used in analysis.

12

I. A panel of twelve college age, non-expert, Caucasian
examiners were each provided with a copy of the filled out
morphological assessment sheet, and shown the pool of eight
‘suspects', including the target. The panel was asked to
eliminate suspects, based on the feature descriptions on the
morphological assessment sheet. The goal of the examination

was to try to locate the target by process of elimination.

II. The panel of twelve examiners were provided with a
blank morphological assessment sheet, and were asked to fill
it out for one specific photograph (the photograph
designated '#1' in the pool of suspects). The purpose of
this was to assess the consistency of descriptions among the

examiners.

l3

Results

Part one of the analysis was that in which the
examiners were asked to eliminate suspects from a pool of
eight based on the morphological assessment sheet. In the
frontal view (Table 1.), all of the examiners were able to
narrow the suspect pool to three suspects or less, without
eliminating the target photograph. Eleven of the twelve
examiners successfully narrowed the pool to one photograph
(the target photo). In the profile view (Table 2.), all of
the examiners were able to narrow the suspect pool to two or
less. There were no false negatives (the target was not
eliminated). Ten of the twelve examiners were able to
narrow the population to one photograph (the target photo).

Part two tested the level of agreement among examiner
responses. The criterion for high agreement was considered
to be agreement of ten or more examiners. In the frontal
view analysis (Table 3), a: high level of agreement was
achieved in nine of twenty-five categories (the category
‘Other’ was excluded). The categories which had high
agreement were as follows: forehead height, hairline
recession shape, degree of baldness, eyebrow thickness,
nose bridge breadth, nose tip width, nostril visibility,

upper lip thickness, and gonial eversion. Int the profile

view (Table 4), a high level of agreement was achieved in

14

twelve of twenty-five categories (category ‘Other' was
excluded). The categories which had a high level of
agreement were as follows: hairline recession shape, degree
of baldness, eyebrow thickness, septum tilt, lower lip
thickness, lip eversion, mouth corner, chin projection,

helix roll, presence of Darwin’s point, and ear lobe.

15

Discussion

The primary objective of this project was to develop a
systematic protocol for facial comparison, based on
existing classification lists. The goal was to modify the
list to be useful to law enforcement personnel, as well as
anthropological experts.

The secondary objectives were to determine if the
modified protocol (morphological assessment sheet) could be
used to eliminate suspects from a population, and to assess
the level of agreement among twelve non-expert examiners in
their assessment of a photograph.

Table 5 and 6 depict the number of suspects eliminated
by each examiner in part 1 of the analysis, as well as
which suspects were not eliminated. From this table, one
can conclude that this assessment protocol is a viable tool
for eliminating suspects, and thus narrowing the suspect
population for further analysis.

In part 2 of the analysis, a high level of agreement
was achieved in nine of twenty-five categories in the
frontal View, and twelve of twenty-five in the profile
view. The category of other was eliminated from this
evaluation. In the remaining categories, a lower level of

agreement (nine examiners agreeing or less) was achieved.

16

The most dominant reason for a lower level of
agreement in the other feature categories is most likely
the use of non-expert examiners. Many of the terms can be
taken as subjective; an expert examiner has the training
and knowledge to critically regard features and assess them
in a clinical manner, while a non-expert must rely in part
on his own perceptions.

It is well recognized that it is easier for people to
characterize features of individuals of the same race. In
order to develop meaningful controls, homogeneous ethnicity
was chosen for both the suspect and examiners. In real
criminal justice situations, it is acknowledged that
suspects, targets, and examiners will differ in ethnicity.
While this poses little or no limitation for expert
examiners, lay persons may not have the level of
sophistication to completely overcome this.

Another limitation in this study is the level of
quality of the photographs. As a preliminary study it was
necessary to use professional quality photographs to
determine if the analysis would be feasible, without the
complication of poor quality surveillance video. It must
be noted that in practice, the angulation and photo quality
will not likely be as ideal as staged professional

photographs, and the level of success obtained may differ.

l7

Conclusions and Recommendations

The modification of current anthropological
classifications into a comparative protocol for use by law
enforcement personnel is in and of itself a major benefit
to the law enforcement community. Despite the limitations
(homogeneity,' quality of media, small populations of
suspects and examiners) of this preliminary work, the data
generated from this study indicate that this protocol can
be very valuable in helping non experts eliminate majority
of suspects from a pool, and thus aid in identification.

Future work in this study should include expanding on
the developmental objectives of this study. The
morphological assessment sheets could Ina adapted to
computer. This will allow' for computer generated
schematics of features in the supplemental guide.
Moreover, the supplemental guide could be incorporated into
the morphological assessment sheet for ease of use.
Results could be tabulated from the program allowing for
printout or simple electronic transfer.

The experiments should be expanded and repeated, using
larger cross-sections of both examiners and suspects.
Eliminating suspects from increasingly larger populations

will yield a better understanding of how effective a tool

18

the morphological assessment sheet is for narrowing a
population. In addition, this will allow for evaluation of
whether high agreement is achieved consistently among
certain features. Results from a larger population will
allow more accurate inferences to be drawn about the
effectiveness of each particular feature, which will in
turn allow for refining of the classifications and sub-
classifications of the features included on the
morphological assessment sheet.

Further study, which has been suggested by Vanezis et
al (1996) is to evaluate the concurrence of certain
features. It is suggested that the knowledge of how
frequently certain features exist with others would prove
useful in determining the significance of findings. For
example, a snub nose exists most frequently with a concave
profile; therefore, would it be significant if the person
in the video still frame possessed. a snub nose with a
convex profile? One must be cautious with this and not
attempt to assign statistical probability to the features,
as the morphological assessment is qualitative in nature.

Another area of interest that should.kxa addressed is.
what effect non-expert examiners have on this analysis.
There are several issues which should be addressed,

including whether the supplemental guide and glossary are

19

comprehensive enough to allow the examiner complete
understanding of the feature descriptions. The most
straightforward approach to testing this is to have two
panels of examiners use the morphological assessment sheets
in an analysis, Iwith one panel using the supplemental
guide, and the other panel performing the analysis without
the guide, as a control.

The second issue to be addressed with non-expert
examiners is whether heterogeneity of ethnicity has any
effect (n1 the analytical outcome. This ‘would. ascertain
whether a non-expert examiner possesses the level of
sophistication. to accurately assess individuals of races
differing from his own. The method for testing this would
be to have two panels of examiners(each panel of a
different race), and two pools of suspects (one of each
race the same as the examiners). Have both panels of
examiners evaluate the suspect pools of their own race, as
well as the other to determine if there is a discrepancy in
the results.

After it: is known whether heterogeneity between
examiners and suspects has an effect on the analytical
outcome, it must be evaluated as well if the morphological
assessment sheets are appropriate for races other than

Caucasian. This may be assessed with one panel of

20

examiners evaluating several pools of suspects of various
races. This would most appropriately be carried out by
experts who are familiar with characteristics of different
ethnic groups, and could provide valuable insight on which
characteristics should be eliminated, as well as
characteristics which would provide a more encompassing

description of a particular race.

21

Appendices

22

Appendix A

Tables

23

 

Table 1:

Remaining suspects after frontal view analysis.

 

Examiners

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 2 3 4 5 6 7 8 10 11 12
m «' «' J’ «' J’ J' J’ J’ J' J’ (I
4.)
u
o
o.
m
s
m
J’
/'
Table Remaining suspects after profile view analysis.
Examiners
1 2 3 4 5 6 7 8 10 11 12
./
m
4.)
o
o
a.
(I)
s
(I)
J’ J’ J' J’ J’ J' J’ J’ J' I' '/

 

 

 

 

 

 

 

 

 

 

 

 

 

 

24

 

 

Table 3.

Feature

Form
Elliptical
Round

Oval
Pentagonal
Square
Trapezoid
Forehead Height
Low

Moderate
High

N/A
Forehead Width
Narrow
Moderate
Broad

N/A

Hairline
(Non-receding)
Round
Moderate
Square
Widow’s Peak
N/A
(Receding)
Temporal
Central
Asymmetric
Other

N/A
(Baldness)
Slight
Advanced
Complete

N/A

Brow Ridges
Absent
Visible
Prominent
Eyebrow Thickness
Sparse
Moderate
Thick

Responses

0 HONU‘IOA

HOOD-‘H ##0wa OAQH O

I—‘OON

U

25

Subject evaluation of photograph 1a.

Feature
Shape
Straight
Arched
Wavy
Continuous
Nose
(Bridge Breadth)
Small
Moderate
Large

N/A

(Tip Width)
Narrow
Moderate
Wide

.N/A

(Tip Shape)
Pointed
Rounded
Amorphous
Nostrils
(Visibility)
None
Slight
Visible
(Shape)
Slit
Ellipse
Round

N/A

Alae
Compressed
Moderate
Flaring
Upper Lip
(Thickness)
Thin
Moderate
Thick

N/A

Responses

HOUIO‘

ANNA

HG)!»

OOOH

Table 3 (cont'd).

Feature
Upper Lip
(Notch)
Absent
Wavy
V-Shaped
Asymmetric
Philtrum
Flat
Parallel
Divergent
N/A

Lower Lip
(Thickness)
Thin
Moderate
Thick

N/A

Chin
(Shape)
Round
Cleft

N/A
(Width)
Narrow
Moderate
Wide

N/A
(Projection)
Not apparent
Protruding

Responses

HOMO)

OOKDUJ HHNxI

Ox]

OHCDUJ

.5

Feature
Gonial Eversion
Not pronounced
Slight
Moderate
Everted

Very Everted
Ears

(Size)

Small
Moderate
Large

N/A
(Projection)
Slight
Moderate

Very Everted
N/A

Other Features
Birthmarks
Wrinkles
Scars
Piercings
Facial Hair
Mustache
Beard

Goatee
Sideburns

Responses

l

0000 OU'IUIN 00mm OOOHH

OOOO

Table 4.

Feature
Facial Profile
Jutting
Forward Curve
Vertical
Concave

Low Jutting
Upper Jutting
Forehead Height
Low

Moderate

High

N/A

Hairline
(Non-Receding)
Round
Moderate
Square

N/A

(Receding)
Temporal
Central
Asymmetric
N/A

Baldness
Slight
Advanced
Complete

N/A

Brow Ridges
Absent
Visible
Prominent
Nasion Depression
Slight
Moderate

Deep

Eyebrow Thickness
Sparse
Moderate
Thick

Shape

Straight

Wavy

Arched

Responses

HOOP I-‘OON wmwo OONIUI OOONmb

(Db

27

Subject evaluation of photograph lb.

Feature
Nose
(Profile)
Straight
Concave
Convex
Bridge Height
Small
Moderate
High

Tip Shape
Pointed
Round
Septum Tilt
Down
Moderate

UP

Nostril
(Visibility)
None

Slight
Visible
Upper Lip
(Thickness)
Thin
Moderate
Thick

Very Thick
N/A

Lower Lip
(Thickness)
Thin
Moderate
Thick

Very Thick
N/A
Eversion
None

Slight
Very Everted
Mouth Corner
Downturn
Straight
Upturn

N/A

Responses

‘3

NM

000mm

Table 4

Feature
Prognathism
Absent
Moderate
Pronounced
Chin
(Shape)
Rounded
Pointed
Double
Projection
Negative
Neutral
Protruding
Ear

-(Size)
Small
Moderate
Large

N/A

(Helix)
Flat

Slight Roll
Very Rolled
N/A

(cont’d.).

Responses

b

(3me

OO

28

Feature
(Darwin's Point)
Absent
Present
Lobe

None

Free
Attached
Other Features
Birthmarks
Moles
Wrinkles
Scars
Tattoos
Piercings
Facial Hair
Mustache
Full Beard
Goatee
Sideburns
Other

Responses

OOOOAH

OHOOO

Table 5: Examiner responses in frontal view analysis.

Examiner Number # of Suspects # of Suspects
Eliminated Remainin

 

\I\I\I\I\I\IUI\I\I\I\I\I
u
\
wwwuwwmwuwwu
‘
(I)

HI—‘H
NHommqmmanI-a

Table 6: Examiner responses in profile view analysis.

Examiner Number # of Suspects # of Suspects
Eliminated .

8

8

\IQQQQQQQQQQQ

l
2
3
4
S
6'
.7
8
9
10
11
12

 

29

Appendix B

Figures

30

Morphological Examination
Frontal View

1. Form
Elliptical
Round
Oval
Pentagonal
Square
Trapezoid

Asses?

11. Forehead
A. Height
2. Low
3. Moderate
4. High
5. Not applicable
B. Width
Narrow
Moderate
Broad
Not applicable

PPN?‘

III. Hairline

A. Non-receding
Round
Moderate
Square
Widow’s Peak
Not applicable

9991"."

I'D
ET
on

Temporal
Central
Asymmetric
Other (describe):

PPM"?

5. Not applicable
C. Receding (degree of
baldness)
Slight
Advanced
Complete
Not applicable

:“P’Nt‘

IV. Brows
A. Brow Ridges
1. Absent
2. Visible
3. Prominent
B. Eyebrows
1. Thickness
a. sparse
b. moderate
c. thick
2. Shape
straight
arched

wavy
continuous

9-.“ 9‘!”

V. Nose
A. Bridge Breadth
1. Small
2. Moderate
3. Large
4. Not applicable
B. Tip Width
1.Nmmw
2. Moderate
3. Wide
C. Tip Shape
1. Pointed
2. Rounded
3. Amorphous
D. Nostrils
1. Visibility
a. none
b. slight
c. visible
2. Shape (if visible)
a. slit
b. ellipse
c. round

Figure l. Morphological Assessment Sheet for frontal

analysis.

31

VI.

3. Alae

a. compressed
b. moderate
c. flaring

Mouth

A. Upper Lip

VII.

Figure l

B.

1. Thickness

a. thin

b. moderate

c. thick

d. not applicable
2. Notch

a. absent

b. wavy

c. v-shaped

d. asymmetric
Philtrum
1. Flat (not apparent)
2. Parallel
3. Divergent
4. Not applicable
Lower Lip
1. Thickness

a. thin

b. moderate

c. thick

(I. not applicable

Chin

Shape

1. Round

2. Pointed

3. Cleft

4. Not applicable
Width

1. Narrow

2. Moderate

3. Wide

4. Not applicable

(cont'd.).

VIII.

32

awesome

Projection
1. Not apparent
2. Protruding

Gonial Eversion
Degree of Eversion
Not pronounced
Slight
Moderate
Everted

Very everted

9'94»!°r‘

Ears
Size
1. Small

2. Moderate

3. Large

4. Not applicable
Projection

1. Slight

2. Moderate

3. Very Everted
4. Not applicable

Other Features
Birthmarks (describe):
Moles (describe):
Wrinkles (describe):
Scars (describe):
Tattoos (describe):
Piercings (describe):
Facial Hair

1. Mustache

2. Beard

3. Goatee

4. Sideburns
Other:

Morphological Examination
Profile View

Left

Right (see Appendix A)

Case Number:

 

I. Facial Profile

11.

III.

?

Figure 2. Morphological Assessment

WWPCP?

POW?

Jutting

Forward Curving
Vertical
Concave

Lower Jutting
Upper Jutting

Forehead Height
Low

Moderate

High

Not applicable

Hairline
Non-receding

1. Round

2. Moderate

3. Square

4. Not applicable
Receding (Shape)
1. Temporal

2. Central

3. Asymmetric (describe):

4. Not applicable
Receding (Degree of
Baldness)

1. Slight

2. Advanced

3. Complete

4. Not applicable
Brows

Brow Ridges

1. Absent

2. Visible

3. Prominent

33

. Depression at Nasion

l. Slight
2. Moderate
3. Deep

. Eyebrows

1. Thickness
a. sparse
b. moderate
c. thick

2. Shape
a. straight
b. wavy
c. arched

Nose

. Profile

1. Straight
2. Concave
3. Convex

. Bridge Height

1. Small
2. Moderate
3. High

. Tip Shape

l. Pointed
2. Round

. Septum Tilt

1. Down
2. Moderate
3. Up

. Nostril Visibility

1. None
2. Slight
3. Visible

Sheet for profile analysis.

VI.

A.

VII.

Mouth
Thickness
1. Upper Lip
a. thin
b. moderate
c. very thick
d. Not applicable
2. Lower Lip
a. thin
b. moderate
c. very thick
(1. Not applicable
Eversion
1. None
2. Slight
3. Very Everted

Mouth Corner

1. Downturn

2. Straight

3. Upturn

4. Not applicable
Alveolar Prognathism
1. Absent

2. Moderate

3. Pronounced

Chin

Shape

l. Rounded
2. Pointed
3. Double
Projection

1. Negative
2. Neutral
3. Protruding

VIII. Ear

A.

B.

Size

1. Small

2. Moderate
3. Large

4. Not visible
Helix

1. Flat

Slight Roll

34

Figure 3: Instructional handbook for frontal view analysis.

35

Form - Select the facial form that most closely resembles that of
the subject, based on the schematic drawings below.

 

IIA.

IIB.

IIIA.

D E . F
Forehead (Height) - Select the most appropriate height of the
hairline. If iJ:.is obscured (by hat or hairstyle, for example),

select Not Applicable.

Forehead (Width) - Select the most appropriate width of the
hairline. If it obscured, (by hat or hairstyle, for example),
select Not Applicable.

Hairline (Non-receding) - If the hairline of the subject is not
receding, select the most closely related shape of the hairline,
based on the schematic drawings below. If the hairline of the

subject is receding, select Not applicable.

 

36

IIIB.

IIIC.

Hairline (Receding) - If the hairline of the subject is receding,
select the pattern in which it is receding. Temporal - at the
temples, Central - by the center of the forehead, Asymmetric - no
discernable pattern. If not receding, or if obscured, select Not
applicable.

Hairline (Degree of Baldness) - If the hairline of the subject is
receding, select the degree to which it has receded based on the
schematic drawings below. If the hairline is not receding,

select Not applicable.

 

IVA. Brow Ridges - The brow ridges are the bony protrusions just above
the eyes, at brow level. If they are not apparent, select
Absent. If they are noticeable but not prominent, select
Visible. If they are very prominent, select Protruding.

IVBl. Eyebrows (Thickness) - Evaluate the thickness of the eyebrows.

IVBZ. Eyebrows (Shape) - Based on the schematic drawings below,
evaluate the shape of the eyebrows.

~ V v

N - _ A”,

a & JN A A A A .- V
II

’M.‘ [(V)\ / k I! . .(VX.
A. B. C. D.

VA. Nose (Bridge Breadth) - The bridge is the bony area of the nose
slightly below eye level. Select the appropriate width. If the
bridge is obscured so that you are unable to assess its breadth
(for example, by eyewear), select Not applicable.

VB. Nose (Tip Width) - The tip is the lowermost central area of the
nose (excluding the alae). Select the most appropriate width (do
not consider the flare of the alad. - see Appendix, VD3 for
definition of the alae).

VC. Nose (Tip Shape) - Select the most appropriate shape of the nose
tip.

VDl. Nostrils (Visibility) - Evaluate the degree of visibility of the
nostrils. If they are not apparent, or only very slightly
apparent, select NOne. IIf they are moderately apparent, select

Mbderate. If they are fully visible, select Visible.

37

VD2. Nostrils (Shape) - If the nostrils are visible enough to make a
determination, select the shape nearest to that of the subject.

VD3. Alae - The alae of the nose are the fleshy areas lateral to the
nose tip that house the nostrils. Select from the schematic
drawings below the degree to which the alae of the subject are
flaring (extending outward).

J V

'3'" ‘5" a» "
'3 5” @VP 3, 9’
i) L3 r
‘9, '2” ‘53".

A. B. C.
VIAl. Upper Lip (Thickness) - select the degree of thickness of the
upper lip. If the lip is obscured (from facial hair, for

example), select Not applicable.

VIA2. Upper Lip (Notch) - The notch of the upper lip is the central
area, directly below the nose. Select the most appropriate shape
of the notch from the schematic drawings below. If the upper lip
appears to not have a notch, or it is obscured, select Absent.
The schematic drawing for asymmetric (example D) is one example
only. If the notch is apparent, but does not fall into any of
the other categories, select Asymmetric.

@ © <3 6

VIBl. Lower Lip (Thickness) - Select the degree of thickness of the
lower lip. If the lower lip is obscured, select Not apparent.

VIC. Philtrum - The philtrum is the area on the central line of the
face from the base of the nose to the top of the upper lip, often

characterized by a central groove. If the area is not obscured,
and there is no visible groove, select Flat. If the area is
obscured, so that a determination is not possible, select Not
applicable.
M

933‘ (9‘ 204/“v ‘x‘tb‘ @ 5""

,5 -—-o g

A. B C.

38

VIIA.

VIIB.

VIIC.

VIIIA

 

IXA.

IXB.

XA-G.

Chin (Shape) - Select the shape of the chin. If the shape is
obscured (by facial hair, for example), select Not applicable.

Chin (Width) - Select the appropriate width of the chin. If the
width is obscured so that a determination may not be made, select
Not applicable.

Projection - Projection is the degree to which the chin
protrudes, or appears to be jutting forward. If the chin is not
notably projected forward, or if it is obscured, select. Not
apparent. If it appears to project forward, select Protruding.

Gonial Eversion - Gonial eversion is the outward flaring of the
most lateral part of the mandible (lower jaw). Based on the
schematic drawings below, select the most appropriate level of
eversion. If the lower jaw is obscured so that this
determination is not possible, select Not applicable.

 

Ears - Evaluate the size of the ears. If they are obscured so
that a determination is not possible, select Not applicable.

Ears (Projection) - Evaluate the degree to which the ears are
projecting from the side of the face.

Other features - Use the topics in A through G as a guide to
describe other distinguishing features on the face.

Location - Define the location of the feature using the
regions depicted in Figure 1 as a frame of reference (e.g.,
mole in upper left corner of L2a). For describing
relations of points on the face, use the glossary of terms

for reference (e.g., the mole is lateral to the nose, and
inferior to the eye).

Size - Describe the size of features only if it is relevant
and adds to the description of the feature. Use subjective terms
only when necessary (e.g., large mole in upper left corner of
L2a, and small mole in center of R4). Use measurements only when
necessary, and in proportion to another feature on the face
(e.g., the distance between the inner corners of the eyes is .4
the distance between the outer corners).

39

XH. Other - Use this section to list any distinctive features on the
face that are not previously in the examination. Note the location
of the feature, and any objective description that will aid in
identification.

L.1

R1

L.22
M L.2b

l
l
R2a ' :
R.2b I

I

I
I
I
.J.._ "___t _ N

L.3

I
m
R3. --I’/
l
I
\V
I

40

Figure 4. Instructional handbook for profile view analysis.

41

View. Select the profile view as it appears in the photo.

5 <

Right Left

IA. Profile - Select the profile that most closely resembles that of
the subject, based on the schematic drawings below.

C” C- C/ c.

A. B. C. D. E. F.
IIA. Forehead Height - Select the most appropriate height of the
hairline. If it is obscured (by hat or hairstyle, for example), select

Not applicable.

IIIA. Hairline (Non-receding) - If the hairline of the subject is not
receding, select the most closely related shape of the hairline, based

on the schematic drawings below. If the hairline of the subject is
receding, select Not applicable.

 

,, LM‘ 5
ED Lib? 2*

42

IIIB. Hairline (Receding) - If the hairline of the subject is receding,
select the pattern in which it is receding. Teaporal - at the temples,
Central - by the center of the forehead, Asymmetric - no discernable
pattern. If not receding or if obscured, select Not applicable.

IIIC. Hairline (Degree of Baldness) - If the hairline of the subject is
receding, select the degree to which. it has receded based on the
schematic drawings below. If the hairline is not receding, select Not
applicable.

     

A. B. C.

IVA. Brow Ridges - The brow ridges are the bony protrusions just above
the eyes, at brow level. If they are not apparent, select
Absent. If they are noticeable but not prominent, select
Visible. If they are very prominent, select Protruding.

IVB. Depression at Nasion - The nasion is the point located on the
median of the face, at approximately eye level. Below are two
profiles superimposed upon each other. The dashed line profile
has a deeper depression at nasion, in relation to the solid line
profile. Evaluate the profile of the subject, and determine the
degree of depression.

.5.

43

IVB.

IVBZ.

VA.

VB.

VC.

VD.

VE.

VIAl.

VIA2.

VIB.

Eyebrows (Thickness) - Evaluate the thickness of the eyebrows.

Eyebrows (Shape) - Based on the schematic drawings below,
evaluate the shape of the eyebrows.

t K
C-

J

U

3

A. B. C.
Nose (Profile) - Based on the drawings below, select the profile
A’ .4 .4,-
C- 0-
5 Q 3
A. B. C.
Nose (Bridge Height) - The bridge of the nose is the bony area

located just below eye level. Select the most appropriate height
of the bridge.

Nose (Tip Shape) - Select the most appropriate shape of the nose
tip.

Septum Tilt - The actual septum of the nose is the fleshy
membrane between the nostrils. Select the most appropriate

direction of tilt (i.e., an 'upturned nose', or 'downturned
nose').

Nostrils (Visibility) - Evaluate the degree to which there is
visibility of the nostrils. If they are not apparent, or only
very slightly apparent, select None. If they are moderately
apparent, select Moderate. If they are fully visible, select
‘Visible.

Thickness (Upper Lip) - Select the degree of thickness of the
upper lip. If the lip is obscured (from facial hair, for

example), select Not applicable.

Thickness (Lower Lip) - Select the degree of thickness of the
lower lip. If the lower lip is obscured, select Not applicable.

Eversion - Eversion (as it pertains to the lips), is flaring
outward, or protruding. Evaluate the degree to which the lips of
the subject evert.

VIC.

VID.

VIIA.

VIIB.

VIIIA

VIIIB

VIIIC

VIIID

Mouth Corner - Evaluate the angle of the mouth corner, if the
subject is in a closed mouth pose. If the subject‘s mouth is
open, or the expression of the subject's face hampers this
evaluation, select Not applicable.

Alveolar Prognathism - Alveolar prognathism is projection of the

upper jaw (overbite). Evaluate the degree of alveolar
prognathism of the subject.

Chin Shape - Evaluate the shape of the chin.

Chin Projection - Evaluate the degree to which the chin is
projecting, or jutting forward. If it appears to have an average
level of projection, select Neutral. If less, then Negative. If

the subject's chin appears to project more than averate, select
Protruding.

Ear (Size) - Select the size of the ear. If it is obscured so
that a judgement of the size cannot be made, select Not visible.

Ear (Helix) - The helix of the ear is the outside edge of the
upper part of the ear, sometimes slightly rolled. Evaluate, and
select the most appropriate description of the helix.

Ear (Darwin's Point) - Darwin's Point is a 'bump' on the inside
edge of the helix. Refer to the figure below to find the
location wher ethe point may be found. If the ear is obscured to
that the point would not be visible, select Not applicable.

Ear (Lobe) - Note the status of the ear lobe. If it is obscured
so that a judgement is not feasible, select Not applicable.

 

Darwin’s Point

45

XA-G. Other Features - Use the topics in A through G as a guide to
describe other distinguishing features on the face.

Location - Define the location of the feature using the
regions depicted in Figure 1 as a frame of reference (e.g.,
mole in upper left corner of L2a). For describing

relations of points on the face, use the glossary of terms
for reference (e.g., the mole is lateral to the nose, and
inferior to the eye).

Size - Describe the size of features only if it is relevant
and adds to the description of the feature. Use subjective terms
only when necessary (e.g., large mole in upper left corner of
L2a, and small mole in center of R4). Use measurements only when
necessary, and in proportion to another feature on the face
(e.g., the distance between the inner corners of the eyes is .4
the distance between the outer corners).

XH. Other - Use this section to list any distinctive features on the
face that are not previously in the examination. Note the location of
the feature, and any objective description that will aid in
identification.

 

 

 

46

Relations

Inferior - located below another stucture.
Lateral - further away from the central line.
Medial — closer to the center line.

Median - on the center line.

Superior - located above another structure.

Glossary of Terms

Alae (as pertains to the nose) - Fleshy areas lateral to the tip of the
nose which house the nostrils.

Divergent - moving apart from a common point.
Flare - to expand or open outward.

Gonial eversion - the flaring of the mot lateral point on the lower jaw
(approximately below the earlobe.

Notch (as pertains to the mouth) — point on the upper lip, often
characterized by a depression of wavy, or V, shape.

Parallel - two lines at equal distance from one another at every point.
Pentagonal - shaped like a pentagon (having five angles).

Philtrum (as pertains to the mouth) - area on the median of the face, from

the base of the nose to the top of the upper lip, often characterized by a
central groove.

Projection - something that thrusts outward: protuberance.
Temporal - Of, or relating to, the temples of the skull.

Trapezoid - A four-sided structure with two parallel sides.

47

References

Austin-Smith, D; Maples, W: “The reliability of
Skull/Photo Superimposition in Individual Identification”
Journal of Forensic Sciences, JFSCA, vol. 39, no. 2, Mar 1994,
pp. 446-455.

Chai, D.-S; Lan, Y.-W; Tao, C; Gui, R.-J; Mu, Y.-C; Feng,
J.-H; Wang, W.-D; Zhu, J: “A Study on the Standard for
Forensic Anthropologic Identification of Skull-Image
Superimposition,” Journal of Forensic Sciences, JFSCA, vol.
34, no. 6, Nov 1989, pp. 1343-1356.

Farkas, L: Anthropometry of the Head and Face. Raven
Press, Ltd. New York, New York. 1994.

Iscan, M; Helmer, R: Forensic Analysis of the Skull:
craniofacial analysis, reconstruction, and identification.
Wiley-Liss, Inc. New York, New York. 1993.

Krogman, W; Iscan, M: The human skeleton in forensic
medicine. Springfield, Ill., U.S.A.: C.C. Thomas, c1986.

Loh, F.C; Chao, T.C: “Skull and Photographic
Superimposition: A New Aproach Using a Second Party’s
Interpupil Distance to Extrapolate the Magnification Factor,”
Journal of Forensic Sciences, JFSCA, vol. 34, no. 3, May 1989,
pp. 708-713.

McKenna, J.J.L; Jablonski, N.G; Fearnhead, R.W: “A
Method of Matching Skulls with Photographic Portraits Using
Landmarks and Measurements of the Dentition,” JOurnal of

Forensic Sciences, JFSCA, vol. 29, no. 3, July 1984, pp. 787-
797.

Sekharan, P.C: “A Revised Superimposition Technique for
Identification of the Individual from the Skull and
Photograph,” JOurnal of Criminal Law, Criminology, and Police
Science, vol. 62, pp. 107-118, 1971.

Stewart, T.D. Essentials of forensic anthropology,
especially as developed in the United
States. Thomas, Springfield, Ill. 1979.

48

Vanezis, P; Brierley C: “Facial image comparison of
crime suspects using video superimposition,” Science and
Justice, vol. 36, 1996. pp. 27-34.

Vanezis, P; Lu, D; Cockburn, J; Gonzalez, A; McCombe, G;
Trujillo, O; Vanezis, M: “Morphological classification of
facial features in adult Caucasian males based on an
assessment of photographs of 50 subjects, JOurnal of Forensic
Sciences, JFSCA, vol. 41, no. 5, September 1996. Pp. 786-791

Yoshino, M; Seta, S: “Personal Identification of the Human

Skull: Superimposition and Radiographic Techniques,” Forensic
Science Review, 1: 23; 1989.

49

nICHIcAN STATE UNIV. LIBRARIES
llllllllllllllllllllllllllllllllllll
31293020741694