u) TECHNICAL REPORT AD NATICK/TR-91/004 · 9 SPONSORING MONTORING AGFV.CS' NAMEIS) AND ADDRESS(Eý. 10. SPONSORING MONITORING AGENCY REPORT NUMBER 1l1..SU.FPpEK'iN1f~l . r N.L CI

u) TECHNICAL REPORT ADNATICK/TR-91/004

N eA COMPUTER PROGRAM FORN •MEASURING FIBERS WITH THE

ZEISS CSM 9500 SCANNING ELECTRON MICROSCOPE

Deborah Sakelakos

Samuel H. Cohen DTICNovember17 199 t L.L 1- r_-

November 1990 DEClIL1990

Final Report ý0June 1990 to September 1990

Approved for public release, distribution unlimited

US ARMY NATICK RESEARCH, DEVELOPMENT ANDENGINEERING CENTERNATICK, MA 01760-5000

SOLDIER SCIENCE DIRECTORATE

ILU

DISCLAIMERS

The findings contained in this report are notto be construed as an official Department of theArmy position unless so designated by otherauthorized documents.

Citation of trade names in this report does notconstitute an official endorsement or approvalof the use of such items.

DESTRUCTION NOTICE

For Classified Documents:

Follow the procedures in DoD 5200.22-M,InduStrial Security Manual, Section 11-19 orDoD 5200.1-R, Information Security ProgramRegulation, Chapter IX.

For Unclassified/Limited Distribution Documents:

Destroy by any method that prevents disclosureof contents or recorsi-ruction of the document.

form AoprovedREPORT DOCUMENTATION PAGE oMB No 0704-0188

b. 0ý, -i -- e orý,,htas-mo

00 "~~o,- Or-ao.,?c"1 na Reoe,1j Q15 Jtfierso,

1. AGENCY USE ONO~ ted.i Diant; 2 REPORT DATE 3 REPORT TYPE AND DATES COVERED

1 5 P nal- .Timi 1QQ() +-n~ qinn QQ4 TITLE AND SUBTITLE 5 FUNDING NUMBERS

A Comrputer Progjran for Measuring~ Fibers with the FE: 612786Zeiss CSM 950 Scanning Electron Microscope

6 0,UTihOF, -I.R 1L162766AH91BIC0oDeborah Sakelakos, and Samiuel H. Cohen9BCO

S7PERF~ORMING ORGANIZATION NAME(S. AND ADDRESS(ES; 8 P[RFORMiNG ORGANIZATION

IU.S. Anrmy Nati ck Retah _rn Ct REPORT NUMBER

Kansas StrieetATI!4: SI'RNC-YSENTC/R9/0Natick, Massachusetts 01760-5020 AJc/T-104

9 SPONSORING MONTORING AGFV.CS' NAMEIS) AND ADDRESS(Eý. 10. SPONSORING MONITORING

AGENCY REPORT NUMBER

1l1..SU.FPpEK'iN1f~l . r N.L CI

Deborah Sake-Lakos was a pdcir-QipdiL iii ti-ie FI'90 UJ.S. Azii-ity Hg CO1ShMathematics Faculty Program ftuicded by the Army Researchl office.

12, DISTRibc I N AVAILAhli- 'S T17 ME I. 12r, D:iTR:13ý:TION COD[

Approved for Public Release, Distribution~ unlimited

13 ABSTRACT (tA',3

17A c=upulter program. designed for the stereological measurement of fibers, waswritten for use with a ZEISS CSM 950 Scanning Electron Microscope (SE2M). T1hisprogjram utilized the ZEISS' build-uin Kontron Image Analyzer to access Surenentsof a variety of paramteters. for fiber analysis of iimges frcin a video cam~era, lightmicroscope or SE2A.

14~ SUJBj!(T TERM. N NM31

, 0 P PACI'

Ccmrputer program, ZEISS L-.12 950, Kontron 1mage Analyzer 26Scanniing Electron Microscope, Fibexs, Images, Parameters 16 FF, ci CO D.

I SE(J CD' C. ASSIF CATION" lb SE CURT' CS ýA5 IFCATIOI, 1S SE. -7 cLA S 43; C! Ti0. 20LI1.`07471N OFABST~t'A?OF R[PDP- 01ý lwr

4 P0(1 0' AB'T4rAC

J UNCLASSIFIE UNCLASSIFIEDL ...jNrASSIEIFT)

._ _, _ _. _ _ _ __.._ _ __-_.._ _ _ _ _ _ __' ,z , . .. .. . . .. . l - S, .m ~ ~ •Z. r T.b r. e•rapThnT Th1ThWJ %p~t 1(•

CENERAL INSTRUCTIONS FOR COMPLETING SF 298

-:ko Reocort Documentation Dage :PDO) .- ý.secl n anniouncing jr'd cataloging reports. it is important.That this information be consistent with the rest of the report. particularly the cover and title page.'is-ruczions for fillba !r zacn oiock of tre form follow. it is r-oortant -to stay within the lines to meet*ptical scanning requirements.

i3!ockl 1, gency Use Or'ly (Leave blark). Block( 12a. Dlsmrýb'o, ;Availability Statement.-'enotes pubjic a-a;ia~ility oriimitations. Cite any

3!ock 2. Recort Date. %ýii ou~b-ca:'on date availability to the oublic. Enter additional~'cldingdaymont, an yea, - a'aoe (eg T"itations oi( spec;al markings in all capitals (e.g.j . ~~n 88). Must cite at least -he ; *EJ EL TR)

3.ock 3. -. --e of Report ar-o Dy;es Cove'ec.~:ae weter eprtsnt.- ~n!, tcDOD ½:e DoDD 5230 24, "Distribution

ipp~icatole, enter ircius~ve re00rt Do u .r. ~aemrts o.Tchia.Un 87 -30Jun 88). Dcmns

DJOE - Reatuhorit~es.3orck 4. -iieard~ub:&;e. A!.-?:: s :~en 'rom NASA ý_ee r~adbook N143 2200.2.m-eaigu anPopeenorao.W'nart of the reoort that oro.ides the -ost N'TI1S ea~Cve blank.

",eoort is prepared in more than one voiurme, ________________

Block 12b. -!stribution Code.I .nclude subtitle for the sceci'c odiume. On O Leepa.casiiid dcumntsentr w e assicaonDOE E- ?.er DOE d~stributzon categories

in arntese. -~.;.or,*1t- S'ardard Distribution forBlocS. undng umbrs. ~ .2c~de ontactjrCassified Scientific and Technical

and grant numbers; may nrclude orogcm AS -jv Oanelement number(s), projec-, nurberWs, NAA -,.,eblrnumber(s), and work un~t niumber(s). ---e the IiTIS - Le~ave ln.following !abels:

C Contract RR - Do~eci Block13l. Abs-ract. ýnclude abrief (MaximumG -Grant rA - ask 2 -Gv.rds) ýactuaý s-,-n~ary of thie mrostPE -Program *NVU - Work Unit s:gn~i FK.2r't . -j:rrmaton ,conta~ned in tne report-

Element Accession No

Block 6. Author(s), Name(s) of person~s) 3iock 14. S:cern-s. <eVywords or phrasesresponsible for writing the report, pc--;ormn~g c ent!,n~g a ý, o.bects in thýe report.the research, or credited with the -ontent of thereport. If editor or compi~er, th's should followthe name(s). Block 15S. Number of Pages. Enter the total

niýmrber of pages.Block 7. Performing Organization Namre(s) andAddress(es). Self-explanatory.

iBlock 16. Dr'ce Code. Enter appropriate priceBlock 8. Performing Organizatorn Reoori code (NTIS only).Number. Enter the unique alphanumeric reportnumber(s) assigned by the organization Blocks 17. - 19. Security Classifications. Self-

perfrmig te rport AgncyNam~s) explanatory En~ter U.S. Security Classification in

Block 9. Sponsoring/Monitoring~gnVNres accordance with U S. Security Regulations (i.e.,and Address(es). Self-explanatory. U%~CLASSlFIED). if form contains classified

Bloc 10 Spnsoing/onioricl~e~cinformation, stamp classification on the top andBlok 0. posoingMoitrin Aeny bottom of the page.

Report Number (if known)

Block 11. Supplementary Notes. Enter Block 20. Limitation of Abstract. This block mustinformation not included elsewhere such as: be completed to assign a limitation to thePrepared in cooperation with ... ; Trans. of ... ; To be abstract. Enter either UL (unlimited) or SAR (samepublished in.... When a report is revised, include as rep' ort). An entry in this block is necessary ifa statement whether the new report supersedes the abstract is to be limited. if blank, the abstractor supplements the older report. is assumed to be unlimited.

S-3aca'a co~rn 298 BzcK ':eV '-89)

TABLE OF COOrrMS

LIST OF FIGURES iv

LIST OF TABLES v

PREFACE vii

IN•ODEUCrION 1

ND1•OES 1

RES= 12

CONCLUSIONS 12

REFERENCE 20

-Accesson For

Tr i tributioi__

Availobility Codes

i iivail nd/or.Wt Special

LUST OF FIGURES

1. Drawing of 1i Fibers to be Measured 2

2. Histogram of Area of 11 Fibers 15

3. Histogram of Perimeter of 11 Fibers 16

4. Histogram of Feret X of 11 Fibers 17

5. Histogram of DCircle of 11 Fibers 18

iv

iv

LIST OF TABLES

~rag

1. Image Analysis Progrwn- for Measuring Fibers 13

2. Statistical Data (in Pix ,is) for 11 Fibers Measured 19

V

PREFACE

In this report we describe the individual cmjponents ard seuenRc of

instructions of a cutputer program designed for the sterelogical

measurerent of fibers. Utilizing on-screen menu onnanrds, we were able to

design and execute a step-by-step analysis of a fiber measurement

program. There was same difficulty in writing the program due to a lack

of clear instructions in the operator's manual; however, conversations

with company representatives helped us overcae these problems.

Vii1

C

A CaX fl FpRGRAM FOR • MEAURI FIBER

WrIn THE ZE=SS CSM 950 SCANING ELBCqWN MICROSCOPE

The objective of this report is to describe a computer pzomrn

designed for measuring stereological features of fibers by means of a

Kontron image analysis system, which is built into a Zeiss CSM 950

Scanning Electron Microscope (SEM).

7The prngram was written to penrit interaction between the menu program

displayed on the TV monitor ani the keybo-ard or rouse. This increased

capability penzits the operator to change parameters within the

measureme-nt pnxociwct deperdaing on the type of fi>ar sample.

since research presently being conducted in these laboratories on

ballistic fibers, including biotiechnologically'derived fibers, requires

information about fiber size, shape, etc., developing a menu- Iled

conput•.r program facilitates that purpose. . ,

MEIHODS

The sazple model use, was a drawing of 11 fibers (Fig. 1) which were

visualized with a Dage MfI Model CD 725 video camera with a Fuji C6 X

17.5 TV zoom lens. ahe images fror the IV camera were transferred to a

Zeiss CSM 950 SEM with a built-in Kontron image analysis system. The CSM

950 utilizes a CP/M operating system having a Z80 processor with a

FIGURE 1. Drawing of 11 Fibers to be Measured.The par-ameters measured (in pixels)were Area, Perimeter, Feret X(maximum projected length in thehorizontal direction), and DCircle(area of fibers within a circlemeasuring 100 pixels in diametex).

2

20 ?.yte hard disk, 64 10byte ranzgon aooass x0ry (RAM) and a 16 KQ6yte

video RAM.

The image processing system performs evaluations in a series of

steps. First, the image fram a TV camera, SEM, light iricroscpe or any

other internal or ex>ternal source is digitized and stored. Next, the grey

image is processed for improved contrast and signal/noise ratio. Then,

certain features are extracted, i.e., unwanted or overlapping features are

eliminated, and features such as size, shape, etc., are analyzed.

Finally, t-he derived data are processod statistically and a printout or

photograph is obtained.

There are several major functions, which am .- ouped dependiiug upon

the specific sequence of a stardard measuring pneh.-.. The functions

cu t2 -

1. Input - defines the type of iiipt device and controls image

digitization and storage.

2. Calibrate - controls the scaling factors, i.e., inches, microns,

etc.

3. Enhance - rescales the grey levels, thereby enhancing image

quality. Image noise reduction fi-ters are used.

4. Image Edit - contains functions for the editing of images.

5. Segment - extracts the ilmoge backgrrain- (distinguishes between

objects and background).

6. Multiphase - enables further processing of binary (two grey

levels) and multiphase (more than two grey levels) images.

3

7. Parameters - permits the selection of object-spacific (single

objects in a field of view), or field-specific (area of a visual field

minus the area occupied by objects).

8. Evaluate - evaluates objects and outputs the results.

9. Advanced - perfonrs geometric and arithmetic transformation of

images, i.e., improves the quality of an image using Fast Fourier

Transform methodss.

10. Utilities - has general auxiliary functions.

11. Peripheral - permits control of peripheral devies, remote

control, etc.

All the above functions have subfunctions that are displayed on the

video monitor in the form of a menu. Dialogue with the program, allowing

the operator to assign or delete nunmerical values as ell as to control

all of the system functions, is accomplished by utilizing the keyboard or

a miouse.

The following is the selection sequence for the operator in order to

obtain the desired program.

A. Calibrate

1- .Scle - Xs al I reAsurement data are indicated in the selected unit of

calibration, the most inportant prerequisite for every geometric

reasurement is calibrating the measuring system. The calibration

procedure is carried out according to the instructions in the operator's

manual (Anonymous, 1986).

a. SCRO, the number of the scale factor to be activated, is set at I.

4

B Ealuate

1. Reset - Before starting a new measuring sequence, e.g., after loading

a new program, scue cc~rditions such as data bufffering, identification

rrber and scanning stage can be reset. The system then has the same

status as after initialization.

a. Single (on) - Clears the buffer into which all the data fromn

several fields are accumulated.

b. Class (on) - Clears the data sumned up over several evaluated

fields to provide average values (e.g., nean area percentage).

C. Select Paratcrter

1. Area - Object specific area.

a. NCIS (20) - Number of classes when classification is selected.

ClassificzItion refc-r to a r--oe f .•_asu-emenent parameter values. i.e.

area, perimeter, etc. into which classes (groups) of fibers would fit.

b. MO3DX (1-3) - Presentation and scaling xmode of the X-axis in the

histogram of the results.

c. MODY (1-10) - Presentation and scaling mode of the Y-axis in the


d. Single (on) - selection of a single list.

e. Class (on) - Selection of a classification.

f. Lcw 0.000 - IU&o-r bound of classification.

g. High 5000 - Upper bound of classification.

D. Select Para-meter

1. Fetvt - Feret diameters (maxinum projected lengths) in X and Y

direction.

a. NCLS (20) - Number of classes when classification is selected.

5

b. MODX (1-3) - Pre-sentation and scaling mode of the X-axis in the


c. MODY (1-10) - Presentation and scaling mode of the Y-axis in the


d. Single (on) - Selection of a single list.


f. Low (0.000) - Lower bound of classification.

g. High (200.00) - Higher bound of classification.

h. 14CLY (20) - Number of classes when classification is selected.

i. MDYX (1-3) - Pres.,entation and scaling mode of the X-axis in the


j. MDYY (1-i0) - Presentation and scaling mode of the Y-axis in the


k. LO.Y (0.000) - Lower bound of the classification.

1. HI.Y (1000.0) - Upper bound of the classification.

E. Select Parameter

1. Perim - Object specific perimeter and length.

a. NCLS (20) - Number of classes when classification is selected,

b. DI2DX (1-1) - _resentation aryi scaling rmode of the X-axis in the


c:. MODY (1-10) - Presentation and scaling mode of the Y-axis in the




6

f. Low (0.000) - Lower bound of classification.

g. High (600.0) - Upper bound of classification.

h. NCLT (10) - Number of classes when classification is selected.

i. MDLX (1-3) - Presentation and scaling mode of the X-axis in the


j. YDLY (1-10) - Presentation and scaling mode of the Y-axis in the


k. IOLT (600.0) - Lower bound of classification.

1. HILT (0.000) - Upper bound of classification.

F. Select Parameter

1. DCIRCL - Diameter of area-equivalent circle (area of fibers within a

circle which measures 100 pixels in diameter).

a. NCLS (20) - Number of classes when classification is selected.

b. MODX (1-3) - Presentation and scaling mode of the X-axis in the


c. MWDY (1-10) - Presentation and scaling mode of the Y-axis in the

histogramn of the results.



f. lcw (0.000) - Lower bound of classification.

g. High (100.00) - Upper bound of classification.

G. Select Utilities

1. LAB: label definition - labels are used as markers in the sequence of

a measuring program to which a jump (analogous to a GOl1 function) can be

executed.

a. I!fL#(l) - Identifier of the label, which serves as designation for

a jump.

7

1. TVa4 - Switches the color monitor directly to a TV signal.

a. Online. (on) - Switches the cnline status on or off.

&. Select Utilijty

1. SYNC - The Image Processing Syste can be synchronized externally

(e.g., by using a TV camera) or intexnally. In the internal mode, no TV

input is possible.

a. Internal (off), Internal (on), or External (off).

J. Select Utility

1. Pause - Interrupts run of program.

K. Irnutm

1. TVINP - This function stores:

a. 'IV iput - A TV image iii the i,,izop ir-icaty b- i

b. INP (1) - Image memory into which the inage is to be stored.

L. Choose Seqment

1. DISC2L - This function separates objects from the background, by

setting two thresholds. Either the grey levels inside or those outside

the entered limits are set to grey value 0 (black) and constitute the

background. Deperding on the variable BIUARY, the remaining object points

either keep their original grey levels or are set to white (255).

a. INP (1) - Image meory oontaiining the image to be prooessed.

b. OUT (2) - Image mmiiory into which the discriminated image is

stored.

c. IEVI (0) - Lower discrimination limit.

d. IMV2 (118)- Upper discrimination limit.

8

e. BINARY (on) - The object pixels become 255 (white), i.e. a binary

image is prodaced.

M. Calibrate

1. MFRAME - This function is used to create a rectangular or circular

measuring frame, which is scsmetimeas necessary to correct errors (eliminate

overlapping images) caused by the finite size of the field of view during

the measuremetnt process.

a. INP (2) - Memory containing the image to be processed.

b. WDSX (508) - X-side of the rectangular frame.

c. WDSY (467) Y-side of the rectangular frame.

d. XO (4) - X position of the (rectangular or circular) frame.

e. YO (6) -- Y position of the (rectangular or circular) frame.

f. RAD (100) - Radius of the circular frame. •

N. Evaluate

1. IDE24T - This function serves two purposes: Identification of

discriminated images and elimination of objects following the conditions

set by the measuring frame (function MFRAME, group Calibrate).

a. INP (2) - Diage memory containing the image to be identified.

b. OUT (3) - Memory containingl the ident if ied and frtme-corrected

imaqe.

c. NRG (1) - Controls the meaning of the measuring frame defined by

MFRAME (function group Calibrate). I1 jis determines whether or not the

fiber touching the rectangular or circular frame on the video monitor

is measured. For example, one might not want to measure a fiber

extending beyond the frame's margin.

9

d. 8-OON (on) - Method of iderti fication, on = 8 CONN.

1. MEASURE - This function ey-xeite- the measurement of selected

parameters such as area, perimeter, etc,.

a. OBJ - Must be switched on if cbject-specific parameters are among

the previously selected parameters (AFEA-TflE).

b. INP (3) - lrage memoxry oontaining the discriminated OBJ image.

c. GRWM (1) - Image memiry containing the original grey image.

d. AIJl (9) - Auxiliary image memry.

e. AUX2 (9) - Auxiliary image irrory.

f. CHAN (1) - Channel selection.

g. SPEC (1) - Distance in pixels between consectiv'e lines of the

"li~n grid". The lines are sinply lines of the image memrory, or pixel

addressez with varying X, but constant V coordinates.

h. FIL# (1) - logical flag ind-icating a buffer overflow.

i. REJECT (on) - Interactive rejection of undesired objects,

preceding an automatic measurement.

P. Evaluate

1. CeTCL -4-4-pu o Clazzi Fj ation T hisz tfunction dsplays thle

results of the classification in the form of a list and/or as a histogram.

a. HIS=I (on) - Provides measured data in histogranm forn.

b. HALT (on) - Presents data on the monitor; the snLm histogram can be

vanipulated.

10

c. NFAC (0.000) - Defines the area (in user units) into which the

absolute counts should be normalized. A value of 0.0 gives normalized

absolate frequencies.

0 . Evaluate

1. CUT'SGL - (Outpuit of single list) This function displays the

individual lists of all neasured data.

a. HALT (on) - Displays lists on the data moriitor after each

nnasureIment.

R. Select Utility

1. Pause - This function interrupts the runnirig of a measuring progran.

1i

RESULTS

The cauiter program designed aid written for the stereological analysis

of several fibers (Table 1) was utilized to measure their Area, Perihieter,

Feret X and D Circle. The irputs to the parameters of this program wre as

follows:

Area (high upper bound of classification) - 5000

Feret X (higher bound of classification) - 200

Feret X (upper bound of classification) - 1000

Perimeter (high upper bound of classification) - 600

DCircle (high upper bound of classification) - 100

The printout of the ccrputer program can be seen in Table 1, the plots for

each parameter can be seen in Figures 2 to 5 and the printout of the

statistical data can be seen in Table 2.

QONCLUSIONS

A carputer program was designed and written to measure certain

stereological parameters of fibers and to generate subsequeit statistical

analysis of the results. Me program was develor-ed to be driven by a series

of menu ccmuw-4s, which allowed for direct interaction beteen the image

appearing on the video monitor and the keyboard.

12

TABLE 1. Image Analysis Program for Measuring Fibers Using the Zeiss CST950 SE24 with a Built-in Kontron Image Analyzer.

WMJOR SUB INTDGER VARIABLES SCALEFUNCTIONS FUNCMIONS VALUES VALUES

A. CALIJATE SCALESCNO I

B. EVLIT RESET --

PAIRAM

C. PARAMEiLRS APEANCLS 20 SINGLE LOW 0.000MODX 1 CLASS HIGH 5000.0MODY 1

D. PARAMETERS 7ERETNCrl5 20 SINGLE LOW 0.000MDDX 1 CLASS HIGH 200.0MDDY 1 LO.Y 0.000NCLY 20 1IU.Y 1000.0MDYX 1MDYY 1

E. PAPAMEV:6 PEPIMNC=S 20 SINGLE LOW 0.000M)DX 1 CLASS HIGH 600.0MODY 2 LOLT 600.0NCLT 10 LILT 0.000

MDLY 1

F. PARAMETERS DCIRCLNCLS 20 SINGLE L) 0. 000M)DX 1 CLASS HICG 100.0MODY 1

G. U17ILITIES 1ABILB• 1

H. INPUr WVON

1. UrILITIE SYNC(Continued)

13

TABLE 1. Image Analysis Program for Measuring Fibers Using the Zeiss CSM950 SE4 with a Built-in Kontron Image Analyzer. (Continued)

MAJOR SUB INfEGER VARIABLES SCALEFUNCTIONS FUNCTIONS VALUES VALUES

J. UITILITITS PAUSE

K. INPJT TVINPINP 1

L. SEGMENT DISC2L BINARYLEVi 0

LEVl 01EV2 118

M. CALIBRATE NFRAMEfINP 2WDSX 508

WEMSY 467XO 4YO 6RAD 100

N. EVALUATE IDENTINP 2 8-ONN-our 3MAUG 1

0. EVTdLATE MEASURINP 3GRIN 1AUXI 9 OB"AUX2 9

SPAC 1FLG " REJECT

P. EVAIUATE OUTcLS__T NFAC 0.000

Q. EVALUATE WTSGL

HALT

R. UrIITIES PAUSE

TOTAL PiF M LEGfli 340 BYTES

14

CAMi'r 11ABS. FREWNCY MD. FLOW 0

Ok7PTLOW 0

10.0 ClA&SL 20%-IL CLASSE

25% 7MEDIAN 9

8.00 75% 1195% 18OM3JL250.0

6.00 L. BOUND0.000

U. BOUND5000.

4.00 oTNTf YM

1142.MAX1MJM

4254.

2.00 - EAI 2177.

0.00

ARE 901.3

0.00 1000 2000 3000 4000 5000

jIGURE 2. Ristogran of Area Measurenlents fron Fiqy1re 1.The X axis values are in pixels.The Y axIs represents the frequency disti ibution.

15

COLUM 11AB. FPXDU4CY M. Flow 0

0

10.0 CLASSE 20I %-IL C~hSSES

8.00 75% 1195% 19

HOWL30.00

6.00 L. BOUND/ 0.000

U. BOUND600.0

4.00 - MINT_165 .7

,4Afl2MJM

:)47.32.00 -KA

289.4SD

108.20.00 PE...-IOM0

0.00 200.0 400.0 600.00

FIGURE 3. Histogram of Perimeter Measure•ents fram Figure 1.The X axis values are in pixels.•The Y axis represents the frequency distribution.

16

aoums 11ABS. F CY UND. FLOW 0

OVERFLOW 010.0 - CLASSES 20

%-IL CLASSE25% 5

MDIAN 68.00 75% 11

95% 20MDDL

20.0006.00 - L. BOUND

0.000U. BOUND

200.04.00 M l

MAXIMIJM

191.02.00 ?TAN

77.91

____ ____ ____ ____ ____ ___48.38

0.00 •= . . . . . . . --. . . . .. FERET X

0.000 100.0 200.0

FIGURE 4. Histogranr of Feret X Meamurerents frnm Figure-. IThe X axis values are in pixels.The Y axis represents the frequency distribution.

17

0OUNTS 11AES. M2LENCY UND. FLOW 0(•RIUMID 010.0 o-ASSE 20

%-IL CLASSES25% 9

MEDIAN 118.00 75% 12

95% 15MDMX1L

5.0006.00 L. BOUND

0.000U. BOUND

100.00

4.00 MI030 .3

73.60

0.00ME

0.000 20.00 40.00 60.00 80.00 100.0

FIGURE 5. Histogram of an Area Equivalent to a Circle(called a DCircle) within Figure 1.

The X axis values are in pixels.The Y axis represents the frequency distribution.

18

TABLE 2. Statistical Data (in Pixels) for 11 Fibers Measured

.cuAREA PERIM .u- R

1 311i.0 413.4 137.0 62.942 2134.0 270.0 53.0 52.133 2512.0 318.4 58.0 56.554 1731.0 250.8 103.0 46.955 2378.0 302.6 52.0 55.036 1691.0 226.0 41.0 46.407 2196.0 269.5 44.0 52.888 1142.0 169.7 68.0 38.139 12).7.0 196.7 79.0 39.36

10 1582.0 219.2 31.0 44.88Ii 4254.0 547.3 191.0 73.60

19

REFERENCE

Anoynois, 1986. "Y•ntron SEi-IPS OperatOr's manual," Volnme II, Release

4.4, 410 pages.

This document reports research undertaken at the

US Army Natick Research, Development and En~in~ejipgCenter and has been assigned No. NATICK/TR-/K I/[gI/in the series of reports approved for publication.

20