(U) ARMY MISSILE COMMAND REDSTONE …AID-A1i51 711 RAYLEIGH CRITERIA RESOLUTION OF OPTICAL...
Transcript of (U) ARMY MISSILE COMMAND REDSTONE …AID-A1i51 711 RAYLEIGH CRITERIA RESOLUTION OF OPTICAL...
AID-A1i51 711 RAYLEIGH CRITERIA RESOLUTION OF OPTICAL CORRELATIONS 1/1(U) ARMY MISSILE COMMAND REDSTONE ARSENAL AL RESEARCHDIRECTORATE D A GREGORY SEP 84 AMSMI/RR-84-i@-TR
UNCLASSIFIED SBI-AD-E950 661 F/fl V/8 N
L3.
Z 1 1E 2O
111
1,1" "
MICROCOPY RESOLUTION TEST CHARTNATIONAL BUREAU OF STANDARDS- 963-A
.o . -
* U TECHNICAL REPORT RR-84-10
RAYLEIGH CRITERIA RESOLUTION OF OPTICAL CORRELATIONS
Don A. GregoryResearch DirectorateUS Army Missile Laboratory
SEPTEMBER 1984
meFcedetoprje ^mAreenal, Alabmma 35S*9
Approved for public release; distribution is unlimited.OTIC
OLECTE
MR2 11985
515 057
Wi F0 ION, I N 61 PWWM*O EITION MAY 1W UME
Ia
DISPOSITION INSTRUCTIONS
WHEN THIS REPORT 1 NO LONGER NEEDED. DEPARTMENT OF THEARMY ORGANIZATIONS WILL DESTROY IT IN ACCORDANCE WITH
THE PROCEDURES GIVEN IN AR 304.
t.
DISCLAIMER
THE FINDINGS IN THIS REPORT ARE NOT TO BE CONSTRUED AS ANOFFICIAL DEPARTMENT OF THE ARMY POSITION UNLESS SO DESIG-NATED BY OTHER AUTHORIZED DOCUMENTS.
TRADE NAMES
UE OF TRADE NAMES OR MANUFACTURERS IN THIS REPORT DOES
NOT CONSTITUTE AN OFFICIAL INDORSEMENT OR APPROVAL OF
THE UE OF SUCH COMMERCIAL HARDWARE OR SOFTWARE.
* . . * - * - .*
- . - .- , .- - .. .- -- V .- '- - -
Unclassf ledSECURITY CLASSIFICATION OF THIS PAGE (fhmm, Data Entered) _ _ _ _
REPORT DOCUMENTATION PAGE READ INSTRUCTIONSBEFORE COMPLETING FORM
T REPORT NUMBER 12 GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER
RR-84-I0 qD.,Ar "1/ __._
4. TITLE (and Subtitle) S. TYPE OF REPORT & PERIOD COVERED -
Rayleigh Criteria Resolution of Optical TechnicalCorrelations
6. PERFORMING ORG. REPORT NUMBER
7. AUTHOR(W) 8. CONTRACT OR GRANT NUMBER(s)
Don )L. Gregory1
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT. TASKAREA & WORK UNIT NUMBERS""
Commander, US Army Missile Command A U NM
Attn: AMSMI-RRRedstone Arsenal, AL 35898-5248
II. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE
Commander, US Army Missile CommandATTN: AMSMI-RPT IS. NUMBER OF PAGES
Rddstone Arsenal, AL 35898-524] ..
14. MONITORING AGENCY NAME & ADDRESS(Qi different from Controlling Office) IS. SECURITY CLASS. (of this report)
ISa. DECLASSIFICATION/DOWNGRADINGSCHEDULE
16. DISTRIBUTION STATEMENT (of this Report)
XpItoved fof public release; unlimited distribution.
17. DISTRIBUTION STATEMENT (of the abstract entered In Block 20, It different from Report)
IS. SUPPLEMENTARY NOTES
I9. KEY WORDS (Continue on reverse aide it neceeay 'ad Identify by block number)
20. ABrRACT IC0f0e - re mm e b n.ei*y84Fad Wadit by block number)
This report describes an investigation into determining the resolv-
ability between correlations produced by Fourier transform matched filtersmade of very similar scenes. Knowledge of this type is needed to accuratelypredict the total number of matched filters needed to completely identifya scene or object regardless of its rotational position. Similar experimentsmay be done for scale and tilt as the variable. This report introduces theRayleigh criteria as the standard by which distinguishability may be measuredin a Vander we
DO I 07 14Tn Emno or FOV 65 IS OBSOLETE
SECURITY CLASSIFICATION OF THIS PAGE (lhe Data Entered)
- _- . . - . . . , v - . , .. ,. .,-.
SECURITY CLASSIFICATION OF THIS PAGE(Whan Data Entw*n
SECURITY CLASSIFICATION OF THIS PAOE(Then Data Battered)
* .n..-.4
6
ACKNOWLEDGEMENT
The author wishes to acknowledge the able assistance of James C.Thompson, currently of Auburn Universit', for his devleopment of the plottingroutines used in this report.
LAccesson ForNT TS G;I
Availa~bility CodesAvail and/or
Dist Special
iii (iv blank)
TABLE OF CONTENTS
PAGE
I . INTRODUCTION ...... ..... *... .... ... . 1 .... ... . *
III* EXPERIMENTAL RESULTS .......... .**.* .. o.o.o... o.o*...*.... .o 4
IV OCLSO.........................
I. INTRODUCTION
Considerable interest has been generated in using Fourier transform
matched filters for object discrimination and tracking. In order to recognize
several objects or several orientations of the same object many matched
filters may be needed.' Interest has been shown in storing several of these
filters in a small area [1,21. This has been done by multiply exposing the
photographic plate or by using arrays produced by holographic elements orother means [3,4]. Arguments are presented here which investigate how closely
packed these filters can be and still have good resolution between them. Inthis instance close packing refers to double-exposing a photographic plate toan input scene at two slightly different rotation positions. Classic line-
shapes are used to model examples of experimental results and some conclusionsare drawn using the Rayleigh criteria as a guideline.
".1o
- -. . - . ..
II. THEORY
The Gaussian intensity distribution has the form
2 2II = Io e-2r /w (.)
where r is the radius from the center of the distribution and 2w is the full
line width at I =(I/e2 )lo. If another distribution of the same form is
located a small distance 61 away
12 - Io e-2(r-61) 2/w2 (2)
then the sum of these distributions is
I 11 + 12 10 [e-2r2/w2 + e -2(r-61)2/w2] (3)
LetI (r=O) . ,:
Pl I(r= c1 /2) (4)
Substituting from the above gives
(el /2w2 e3612/2w) ((5)i
P, 1/2 e+ e 5 ii
P1 is thus the ratio of the peak at r = 0 to the height of the valley betweenthe two distributions. Solving for 61 using the fact that the second term ismuch smaller than the first gives
61 w [2 Ln (2p) ]2. (6)
2
..........................................-.-.~...-....--.....-..............-.. . ..- ". ' " " " ' " " ' ' " -" " ' -" - " -" - " ' -' ." " " - ." ' ' " - ' " -" - , " " " "".". ." ".. . " 4 " " " " " ." "
' ': ' :
Using the Rayleigh criteria as a rough guideline, P1 = 1.23 and substitutingthis into Eq. (6) yields
61 1.34 w . (7)
The same sort of analysis may be applied to a Lorentzian distribution havingthe form
A + A
2 2 2 2 (8)r + (a/2) (r-6 2) + (/2)
where A is a constant equal to 10(o/2)2 , r is the radius and a the full
linewidth at half the maximum intensity. This produces a 62 of
62 o/2 4P2 -3 +416P2(P2-1)+l] 2 (9)
Then for P2 1.23 as a value of the intensity ratio
62= 1.03 a (10)
For a sinc 2 distributionF 12
.21 sinr 2 sin Q (r-6 3)
0 ok0r a(r-6-) 3(I)
where again r is the radius from the center of the distribution and
a = /r(o) (12)
where r(o) is the value of r at the fist zero of intensity. This distribution
yields
3 + sin2 63 (13)
8 sin 2 (a6 3/2)
which must be solved graphically or by iteration using P3 = 1.23. For smallvalues of 63 , note that a2 632 > > sin263 '
3* *~ * -* *-.-X*. ...- ,.
V W P. 7 ..
III. EXPERIMENTAL RESULTS
Figure I shows three distributions compared with experimental data taken
using visible (Helium-Neon) real-time optical correlation methods. The inputscene is fed via an RCA television camera (model number TC 1005) into a
Videotek Monitor (model VM-12PR) which is used as the input to a Hughes ..
liquid crystal light valve (the same one used in Ref. 4) which produces the
coherent image that is fed into a real-time Vander Lugt type correlator. - -
Experiments of this type are described in detail elsewhere [5]. All threedistributions have been arbitrarily fit to experimental data at r - 0 and r =
2.940 using w - 2.80, a - 2.07, and a = .775 deg-1 . This gives 61 = 3.80,
62 - 2.13, and 63 - 4.0% Note that these widths should depend upon the
spatial frequency of the input scene as well as other factors and thus serveonly as examples here. In this particular experiment the input scene was alow resolution black and white aerial photograph of Huntsville, Alabama. Thephotograph was placed on a rotatable stage and matched filters made at the
desired rotation angles. Experimental data in Figure 1 is the correlation
intensity for a filter made at 00 rotation of the input scene. After develop-
ment of the film plate (Kodak 649F) it was replaced in the correlator and the
input scene rotated slowly as the corresponding correlation intensity wasmeasured using an RCA CCD camera fed into a television monitor then digitizedby a Colorado Video Analyzer (model 321). This signal was then recorded by a
Hewlett-Packard model 680 strip chart recorder. Figure 2 is representative of
actual data taken as the input scene was rotated through the position (0
rotation) where the matched filter was made.
In order to investigate multiply stored filters, a single plate was
exposed twice with different rotation angles of the input scene. After deve-
lopment using standard techniques for Kodak 649F plates, the filter was rein-
serted into the correlator and data taken as before. Figure 3 shows the
actual data and a plot of Equation 8 using the Rayleigh criteria. The expo-
sures were made using the same criteria for 62 of 2.130 of scene rotation.
The agreement is reasonably good. Some variation is expected due to the dif-
ficulty in obtaining equally intense correlations for the two filters. This
depends upon the exposure times used, film response, and several other
variables. In the theoretical analysis the calculations could easily have
been done using distributions having different peak heights but then the value
of 6 would depend upon these heights which can't be known with any certainty
in advance of doing the experiment. In the calculations presented, it was
also assumed that individual correlation intensities should be added directly
rather than adding the corresponding fields then squaring for the intensity.
This has been justified experimentally by storing one filter made of 00 scene
rotation and another filter at 900 scene rotation then adding the resulting
correlation intensities algebraically. This distribution was then compared
with data similar to Figure 3. The results were found to be essentially
equal. This indicates that there is little if any phase addition contribution
to the sum of the two correlation intensities.
* - -.. .....
"...'' .' *.. .',s.e, '. .... .... . . . ............................................. .......... .. . . .,..
IV. CONCLUSIONS
In this brief report a method of analyzing closely spaced optical matchedfilters has been presented along with experimental data which tends to supportthe findings. The use of the Rayleigh criteria may not be strictly correct in
that this criteria is for a Bessel function intensity distribution only but it
serves as an initial guideline [6]. Experimental data seems to suggest that theseparation criteria will become more stringent as laboratory methods improve.
5
.1h . d . . .
13
12 '
9..Gaussian
Lorentzian
27" 51nc
6 ~5 ** * * Experimental
2 1 23
SCENE ROTATION (DEGREES)
Figure 1. Correlation intensity vs angle of rotation of input scene
6
wS
0))
to
0 V4
S
-T c
-v4i
V) 0
00
0
Ctlo 0o0
*0
0k'-I
ca
CDl
r-
K:4
ow 4000dfcn
q'a'-W
0 r
44
o4 40
4 q0
04
(SII ~ ~ ~ ~ ~ -AUUI8) ISUN
References
1. P. Meuller, Appi. Opt., Vol. 8, No. 2, 1969 (297).
2. K. Lieb and J. Mendelsohn, Proc. SPIE, Vol. 317, 1981, (148).
3. K. Lieb, R. Bondurant, S. Hsiao, R. Wohiers, and R. Herold, Appi.
Opt., Vol. 17, No. 18, 1978, (2892).
4. H. K. Liu and J. H. Duthie, Appi. Opt., Vol. 21, No. 18, 1982, (3278).
5. J. Smith, B. Guenther, and C. Christensen ,J. Appi. Photo. Engr.,Vol. 5, No. 4, 1979, (236).
6. E. Hecht and A. Zajac, Optics, Addison-Wesley, Reading, Massachusetts,1974, p. 352.
9e..'.
9.•"-
DISTRIBUTION
No. of of
Copies
CommanderUS Army Research OfficeATTN: AMXRO-PH, Dr. R. Lontz 5
P. 0. Box 12211Research Triangle Park, NC 27709
US Army Research and Standardization Group (Europe)ATTN: AMXSN-E-RX, LTC D. R. Reinhard -
Box 65FPO New York 09510
CommanderUS Army Materiel Development and Readiness CommandATTN: Dr. James Bender 1
Dr. Gordon Bushey 15001 Eisenhower AvenueAlexandria, VA 22333
Headquarters, Deaprtment of the ArmyOffice of the DCS for Research, Development & Acquisition
ATTN: DAMA-ARZRoom 3A474, The PentagonWashington, DC 20301
OUSDR&ERoom 3DI079, The Pentagon 1Washington, DC 20301
DirectorDefense Advanced Research Projects Agency1400 Wilson BoulevardArlington, VA 22209
OUSDR&EATTN: Dr. G. GamotaDeputy Assistant for Research (Research in Advanced Technology)Room 3D1067, The PentagonWashington, DC 20301
Director of Defense Research and EngineeringEngineering Technology 1
Washington, DC 20301
::::::
S. ' . - *,.. . ~ % % .. --.. ' S-"
DISTRIBUTION -Continued
DirectorDefense Advanced Research Projects Agency/STOATTN: Commander T. F. Weiner
D. W. Waish1400 Wilson BoulevardArlington, VA 22209
CommanderUS Army Aviation Systems Command12th and Spruce StreetsSt. Louis, MO 63166
DirectorUS Army Air Mobility Research &Development LaboratoryAmes Research Center
Moffett Field, CA 94035
Commander
US Army Electronics Research &Development CommandATTN: AMSEL-TL-T, Dr. Jacobs
DELEW-R, Henry E. SonntagFort Monmouth, NJ 07703
Director
US Army Night Vision LaboratoryATTN: John Johnson
John DelinePeter VanAtta
Fort Belvoir, VA 22060
CommanderUS Army Picatinny Arsenal
Dover, NJ 07801
CommanderUS Army Harry Diamond Laboratories2800 Powder Mill RoadAdelphi, MD 20783
CommanderUS Army Foreign Science and Technology Center
ATTN: W. S. Alcott
Federal Office Building220 7th Street, NECharlottesville, VA 22901
CommanderUS Army Training and Doctrine CommandFort Monroe, VA 22351
DISTRIBUTION - Continued
DirectorBallistic Missile Defense Advanced Technology CenterATTN: ATC-D
ATC-O 1ATC-R 1ATC-T 1
P.O. Box 1500Huntsville, Al 35808
CommanderUS Naval Air Systems CommandMissile Guidance and Control Branch IWashington, DC 20360
Chief of Naval ResearchDepartment of the Navy 1Washington, DC 20301
CommanderUS Naval Air Development Center ,Warminster, PA 18974
Commander, US Naval Ocean Systems CenterCode 6003, Dr. Harper Whitehouse 1San Diego, Ca 92152
Director, Naval Research LaboratoryATTN: Dave Ringwolt 1
Code 5570, T. Gialborinzi 1Washington, DC 20390
Commander, Rome Air Development CenterUS Air ForceATTN: James Wasielewski, IRRCGriffiss Air Force Base, NY 13440
Commander, US Air Force, AFORSR/NEATTN: Dr. J. A. NeffBuilding 410, Bolling Air Force BaseWashington, DC 20332
Commander, US Air Force Avionics LaboratoryATTN: D. Rees 1
W. Schoonover IDr. E. Champaign IDr. J. Ryles 1Gale Urban 1David L. Flannery .
Wright Patterson Air Force Base, OH 45433
!p
DISTRIBUTION- Continued
DirectorDefense Advanced Research Projects Agency/STOATTN: Commander T. F. Weiner 1
D. W. Waish 1
1400 Wilson BoulevardArlington, VA 22209
Commander •
US Army Aviation Systems Command.112th and Spruce StreetsSt. Louis, MO 63166
DirectorUS Army Air Mobility Research & Development Laboratory
Ames Research Center 1
Moffett Field, CA 94035
CommanderUS Army Electronics Research & Development CommandATTN: AMSEL-TL-T, Dr. Jacobs 1.-
DELEW-R, Henry E. Sonntag 1
Fort Monmouth, NJ 07703
DirectorUS Army Night Vision Laboratory
ATTN: John Johnson .
John Deline 1Peter VanAtta 1
Fort Belvoir, VA 22060
Commander 0:
US Army Picatinny Arsenal 1 -
Dover, NJ 07801
CommanderUS Army Harry Diamond Laboratories I
2800 Powder Mill Road
Adelphi, MD 20783
CommanderUS Army Foreign Science and Technology Center
ATTN: W. S. Alcott 1
Federal Office Building220 7th Street, NE .1 -
Charlottesville, VA 22901
Commander " ' '
US Army Training and Doctrine Command 1
Fort Monroe, VA 22351
7' • " " " "
DISTRIBUTION - Concluded
Professor Anil K. JainDepartment of Electrical Engineering 1
University of California, DavisDavis, CA 95616
Terry TurpinDepartment of Defense9800 Savage RoadFort George G. Meade, MD 20755 1
Dr. Stuart A. CollinsElectrical Engineering Department - -
Ohio State Univeristy I
1320 Kennear RoadColumbus, OH 43212
US Army Materiel Systems Analysis ActivityATTN: ANXSY-MPAberdeen Proving Ground, MD 21005 1
US Army Night Vision Laboratory
ATTN: DELNV-L, Dr. R. Buser 1
Ft. Belvoir, VA 22060
Dr. F. T. S. YuPenn State UniveristyDepartment of Electrical Engineering 1
University Park, PA 16802
Dr. William P. BlehaLiquid Crystal Light Valve Devices 1
Hughes Aircraft Company6155 El CaminoCarlsbad, CA 92008
AMCPM-PE-E, John Pettitt 1
-PE 1
AMSMI-LP, Mr. Voigt 1
-0 1
-Y-R, Dr. McCorkle 1
-R, Dr. Rhoades 1
-RN, Jerry Hagood 1-RE, W. Pittman 1
-RD 1
-RG, J. A. McLean 1
-RR, Dr. R. L. Hartman 1
Dr. J. S. Bennett 1
Dr. J. G. Duthie 1
Dr. D. A. Gregory 80
-W =r*.°
DISTRIBUTION -Continued
Commander, AFATL/LMTATTN: Charles WarrenEglin Air Force Base, FL 32544
Environmental Research Institute of Michigan
Radar and Optics DivisionATTN: Dr. A. Kozma 1
Dr. C. C. Aleksoff 1
Juris UpatnieksP. 0. Box 8618
Ann Arbor, MI 41807
IIT Research Institute ..
ATTN: GACIAC 1
10 West 35th Street
Chicago, I1 60616
Dr. J. G. Castle9801 San Gabriel, NE 1
Albuquerque, NM 87111
Commander, Center for Naval Analyses
ATTN: Document Control 1
1401 Wilson BoulevardArlington, VA 22209
Dr. J. W. Goodman
Information Systms LaboratoryDepartment of Electrical Engineering IStanford UniversityStanford, CA 04305
Eric G. Johnson, Jr. 1National Bureau of Standards
325 S. BroadwayBoulder, CO 80302
Dr. nicholas George
The Institute of Optics 1
University of RochesterRochester, NY 14627
Naval Avionics Facility 1
Indianapolis, IN 46218
Dr. David CassasentCarnegie Mellon University 1
Hamerschage Hall, Room 106Pittsburg, PA 15213
%°°-. e e -e--e--'r-
FILMED
4-85
DTlC