FIG. I. Terrain model and framework.

PROF. ED\\'.\RD F. l~uLH.\N*

Fresno State College

Fresno, Calif.

Terrain Model Aerial Photos

College students return to the sand box to experience

salient principles of photogrammetric mapping.

(/1 bstract 01/ page 10--19)

1:-1 ORDER to stimulate interest and to makethe su bject more understandablc, the

author sought to teach a college course inphotogrammetry as a laboratory project. Thepart that seemed to gi \'e the studcn t thcgreatest amount of difficulty was the \'isuali­zation of flight planning and relation of thefocal length of the camera to scale of thefinished photograph. To most students thecamera was simply a black box that \\'aspointcd, triggered, and exposed film to betaken to the drug store for dC\'eloping andprinting. i\lany of the students had ne\'er

AO\\'n in an airplane and, as a rcsult, had noidca of \\'hat thcy would, or should, sce on anaerial photograph.

In order to gi\'c a bettcr concept of whatthe student nccds, let mc re\·iew a little back­ground on thc course as taught at FresnoState Collcgc. I~ngr. 103, Photograillmetry, istaugh t to ci \·il engi necri ng studen ts, not wi ththe idea of maki ng photogram mctrists ou t ofthcm, but to strcngthcn thcir foundation ofkno\\'ledge about sun'eying and mapping.\\'ith this in mind. it seemed that a projectfor mapping a gi\'cn area from start to finish

* Presented at t\nnual COl1\'ention of the .\merican Society of Photogrammetry, \\'ashington, D. c.,:\Iarch 1965.

1046

TERR,\IK :\IODEL ,\ERIAL PHOTO 1047

offered the bcst means of attaining this aim,IlolI'e\'er, obtaining adequate picture co\'er­age \\'ith ground control is usually too difficultand expensi ve,

How to obtain this co\'erage and make it acomplete projcct from planning stage to atopographic map lI'as soh'ed by simulation inthe laboratory, \\'e used a terrain model in asand box \\'ith de\'ices to duplicate aerialphotography, Figure 1 sho\\'s the frame, mod­el, camera track and boom, In this ,,'ay thecamera can be mo\'ed along the track to simu-

computes the picture !:pacing and flight spac­ing for their plotting scale, Each group setsthe camera height and takes its picturcs atthe propcr spaci ng, The Ii 1m is then de\'elopcdin class, con tact pri nts made, and scale mea­surements checked to verify the results, Fromthe contact prints (Figure 2) the co\'crageand o\'crlap \\'as checked, During the pastsemestcr six groups of students \\'ere ill\'olvedand all camc up lI'ith the propcr answcr, Theauthor belic\'es that lI'ith this simulatedproblem thc students ha\'c a greatcr intercst

FIG, 2, COlltact prillts, 'fhc original scale II';\S till, =570 ft, alld the flight height lI'as 85,5 ill,

late the flight of the planc, Thc track can bemo\'ed to parallel positions to give properlyspaced flight lincs, The studcnts arc shownthe model and gi\'en the model scale, "'hich inthis case is t inch equals 20 fcet, The modelis 8 by 8 ft" thus cO\'ering an area 1920 by1920 ft,

Each group of three studcnts is gi\'cn aplotting scale to be uscd in multiplcx compila­tion and told to compute thc flying height ofthe camera abo\'c the a\'eragc terrain ele\'a­tion to give this scale, Thc multiplex has aprincipal distancc of 46 mm, and optimumprojection distance of 360 mm, thus gi\'ingan enlargement factor of 7,84, The camcrahas a focal length of 75 mm,

Consider, for example, a plotting scale of1 in, = 60 ft. The picture scale would thenha\'e to be 7,84X60, or 1 in,=470,4 ft, Thiswould represent 470,4/20 = 23,52 in, on themodel. :\s the negati\'e is 1.625 in, in direc­tion of flight, it would covcr 1.625 X23,52= 38,22 in, on the modcl. Wi th 60% o\'erlap,the net progress is 40%, and the picture spac­ing is 38,22XO,40= 15,28 in, In a similarmanner the flight spacing can be computcdexccpt that the \\'idth of the negative is 2,125in, instead of 1.625, Six pictures are needcd incach flight line,

In a similar manner each group of students

and incen ti \'e to learn these fundamen talsand \'isualize the real problcm, They knowthat their work is easy to check and thecom peti tion betll'een grou ps is ra ther intense,The studcnts are the se\'crest critics of eachother.

For ground control, targets as seen in Fig­ure 3 \\'ere sent out on ~-inch squarc blocks of3,4,5, and 6 inches in height and fastened tothc bottom of the sand box, The blocks \\'crc

P'WF, EO\\'AIW F, KL'LHAX

1048 PHOTOGRAMMETRIC E;\JGINEERING

FIG. 3. A stereoscopic pair of photographs of the terrain model. The original scale was 1 in. =285 fl.

uscd to eliminate settlement of the sand underthe target \\·hen a mcter stick used for Icvelrod \\"as placcd on thcm. Thesc were arbi­trarily sct for a flying height of 90 in. andspaccd ]9 in. apart in direction of Right andon both thc right and left sidc of the pictureso that the con trol \\"as in an ideal posi tion foruse as pass poi n ts. This arbi trary spacing hadits disac!\'anlagc in that it \\"as not in propcrposition for othcr than thc one for \\'hich itwas set. 1 in. = 7S ft. I n searching for anothermethod of anchori ng the can trol poi n ts (theprcscnt mcthod makcs it impossible tochange thc control points \\"ithout disturbingthe model), we ha\"e come up \\"ith the ans\\"erby using 16-penny spikes cut to desired lengthand points sharpened. The spikes can bedri\'en into the bottom of the box, the targetglued to the head \\"ithout disturbing themodel in any \\·ay. In this way thc controlpattcrn can bc changcd for each scale desired.

By orienting the hox Sorth and South, onecdge can sCr\'e as thc 1'- V (N-S)-axis andthe other edge as thc x-x (E- IV)-axis withone corncr as the origi n. Thc grou nd co­ordinates can he mcasured and elevationsdetcrmi ned by COI1\'en tional levels, I n addi­tion to thc cOl1\'enicnce of coordinate axis, byorienting the edges N-S and E- W, thc shad­O\\'S are in natural position, and more realis­tic pictures can be obtained,

J"\s an aid in leveling the multiplex model,the left and righ t pass"poin t can trol poi n ts

\\'ere sct to the same elevation. In this waythe four corncr points on each model \\"cre thcsame cle\"ation and le\"eling in the X and Ydirections was simplifled. The instructor hasto hear in mind that he is not producingmultiplex operators, as cach student willprobahly not ha\"e more than four hours oftime on the plotter. Anything to spced up thcoricntation procedurc is ach-antageous. Oncethe stud en t le\'cls the model and learns therelation of horizontal and \'ertical scale, andthc methods of le\'eling the modcl. he can goon to other models that arc not set for idealconditions.

So much for the modcl and thc taki ng ofthe pictures. From thc negati\cs the studcntsmade enlargemcnts approximatcly 4X to a9X7 in. sizc and dctermincd the scalc of theenlargemcnts, From this scale thc plottingscalc for use \\"ith the skctclllllastcr \\'as de­tcrmincd, Our sketchmasters do not ha\"cauxiliary lenses; therefore, thc cnlargementhas to he kcpt to a minimum so that the dis­tance from picture to map docs not becomccxcessi\'e for loss of dctail. Thc plotting scalcused \\'as 1 in. = 140 feet.

:\ext, each student made radial-linc tem­plets of cach picture using thc so-called"spidcrs." Three of the control points of thcmodel \\'ere plotted in posi tion on thc projec­tion and the rcmaining control points (passpoints) as \\"cll as thc principal point of eachphoto \\"erc located by the radial line plot.

TERRr\ll" l\IODEL AERIAL PI-lOTOS .1049

The location of the pass points as obtainedfrom the radial line plot was then checkedand compared Il'ith the knOIUl co-ordinatesof the points on the model. In this manner thestudent gains confidence and realizes thereliability of the radial line plot, \\'here thepoints do not check out properly, the studentalso learns that there is no room for careless­ness, In this case he repeats the work, but onhis O\l'n time, The author flrmly beliel'es thatthe student has to find out for himself thetolerances allowed; in other \I'ords, the pre­cision necessary for \'arious types of II'0rk,

To take adl'antage of the al-ailable equip­ment and to gi\'e the student as much chanceas possible to usc the \'arious methods ofcompilation, the map \I'as compiled on four

II' hich is the size of the diaposi ti \'es used bythe normal angle projectors, This enables usto make the diaposi ti \'es by con tact pri n ti ngon the glass plates Il'i thou t usi ng a reductionprinter. The first thought that will probabh'come to your mind \I'ill be that by contactpri n ting the horizon tal and \'ertical plotti ngscales \I'ill not be equal, and that distortionwill be present in the diapositi\'es as no cor­rection lenses were used, These are legitimatethough ts, bu t there arc good reasons II' hythese \I'ere disregarded, The lirst reason isthat \I'e do not ha\'e a printer and printers forthis size of film are not a\'ailable, The \'erticalscale is dilTerent from the horizontal scale, butthis is of no real conseq uence as Il'e are notin\'oll'ed in production \I'ork lI'here time is of

:\BSTRACT: Tlte problems of illcreasing stlldwt interest ill pllOtogrammetr}' alldtlte sOllrce of aerial pltotograplty for illstrllctional lise It!l!'e plaglled pllOtogram­metr}' illstr ltctors sillce tI,e COli rses were sta rtel!. The II se of Terra in .Ifodel .1 erialPltotos lilts been tlte allswer to both problems at Freslw Stllte Col/ege, Tlte tectl­niqlles of makillg tlte terrain model alld talzing tlte "aerial" pltotogra,plts areinteresting and instrltctil'e tltemseh'es, in addition to tlte ases marie of tltepllOtograplls dllring tlte coarse of instrllctioll,

different pieces of equipment, Bearing Inmind that each student had ten photos toII'ork, photos 1 and 2 \I'ere compiled on themultiplex; 3, 4, anc! 5 on the ~ketchmaster; 6,7, and 8 on the opaq ue projector; as ti me per­mitted the last two, 9 and 10, on the Kailradial line plotter,

\\'e ha\'e on hand three different types ofOpaque Projectors, Olle is \'ery similar to a\'iell' camera; in fact, \I'e use it as a copycamera, This is used solely to change scalefrorn picture to map, No tilt motion is 1'1'0­

\'ided, The second one is a projector \Vi th atilting table utilizing a Johnson head from aplane table, This projector \I'as designed by astudent as a Senior Project, By using the tilt­ing table the same effect can be obtained asIl'ith a sketchmaster, but much faster as \I'ellas allO\l'ing greater latitude in scale change,The third is a :\Iap-o-Craph Il'hich we ha\'ecOIl\'erted to a transform projector, By locat­ing the axis of tilt of each picture and settingit in line Il'ith the axis of the tilting mirror,both the mirror and the table can be tiltedthus in effect rectifying the print, This adap­tation of the :\ Iap-o-Graph was alEa done by astudent as a Senior Project.

For the Illultiplex compilation, the diaposi­ti \'es are made by the studen ts, The nega­ti\'es, as pre\'iously mentioned, are 2l X Ii in"

the essence, bu tin instructional \I'ork \1' herethe prillciple is more important. The studentscan see that the scale Il'ill be in the ratio ofthe focal lengths of the camera and projector.In our case the ratio of thc \'ertical to hori­zontal scale is 75/46 or 1,63, Our tracing tableis of the older type \I'i th the \'crtical standardgraduated in millimeters: therefore, tablescan be built up for \'ertical settings for anycontour elel'ation at any scale, \\'ith this typeof table, compilation at any scale is possihleIl'i thou t ha \'i ng to \\'orry ahou t the proper setof gears.

As for the presence of distortion, this againis accepted and the studen t made aware ofthis, but again you ha\'e to hear in mind thatstudent is not a trained operator and he cannot percei\'e small distortions because he Il'illprobably ha\'e only about four to six hours onthe multiplex, ..-\11 in all Il'e found no apparentdistortion that could hc percei\'ccl and theresults of the compilation donc by the stu­dents was relati\'ely good-actually Illuchbetter than expected, The relatil'e and abso­lute orien ta tion of the model \I'ere not toodi fficul t as the stud en ts seemed to enjoy theIl'ork on the multiplcx and could \'isualize theeffect of each motion of the projector,

As in the radial line plot, somc of the con­trol poi n ts \I'ere not plotted on the projection

1050 PHOTOGRA\I\IETRIC EKGI:-\I':ERI\'G

so as to check their compilation, Again the\I'ork \I'as yery satisfactory and better thanexpected.

The com pilation of the com plete map wasdone on t\l'O different scales, 1 in. = 140 ft. onthe sketch master and opaque projector, and1 in. = 60 ft. to 1 in. = 120 ft. on the multiplex.1n order to acquaint the students withmethods of reprod uction of maps and scalechange as done commercially, the publicationscale of maps 'I'as set at 1 in. = 200 ft. Thetll'O maps were then photographed on thecopy camera ('I'hich incidentally was de­signed and buil t by a studen t as a Senior Pro­ject) to the common scale of 1 in. = 200 ft.The contact prints of the maps were thenmounted on masonite hacking, the compositeagain photographed at 1: 1 ratio, and thejoin lines opaqued out to [orm one map. Fromthe resulting composite negative the studentmade a film positi"e on cronoAex film fromwhich blue line or color separation scribe coatsheets 'I·ere made.

Time permitting, 8i X 11 in. sheets ofscrihe coat were coated II'ith diazo liquid andusing the rronoAex film positi"e three prints

of each map were made on the ozalid ma­chine. Each student scribed one color and acomposite print in three colors \I·as made,usi ng the \l'ashcoat process. The old \l'ell­known sun frame 'I"as used to make thepri n ts.

In this manner the studen t was madefamiliar with the process used to make a com­plete map. The au thor belie"es that in thismanner, by starting with the Aight planningand finishing "'ith a three color map that hasbeen planned, Aown and photographed, com­piled and reproduced by the student, gi"eshim an idea of the process and planning thatgoes into making of a modem map. Thoughthe student may neyer ha"e to do any ofthese things, he \I·ill know \I"hat to ask forw hen he needs a map made, and not ask forthe impossible or ask for precision not com­patible ,,·ith the results needed. The studentis gi,·en a tool to use if he needs it in laterpractice, Again please bear in mind that \I'eare attempting to gi"e a prospecti,'e ci"ilengi neer a tool to ,,'ork wi th and not tryi ngto transform him into a photogrammetristin six easy lessons.