High Altitude Photography: Aspects and Results · 2017-09-10 · D. GUT J. HOHLE Wild Heerbrugg...
Transcript of High Altitude Photography: Aspects and Results · 2017-09-10 · D. GUT J. HOHLE Wild Heerbrugg...
D. GUTJ. HOHLE
Wild Heerbrugg LimitedHeerbrugg, Switzerland
High Altitude Photography:Aspects and ResultsLarge area coverage, combined with the use of highresolution cameras and precision stereoplotters, provides foreconomical mapping procedures.
INTRODUCTION
I N PRACTICAL PHOTOGRAMMETRY a singlephotograph is used to cover as large an
area as possible while maintaining recognition of detail and geometric accuracy necessary for topographic mapping. Small jet aircraft enable the photogrammetrist to attainimage scales of up to 1: 150 000. Both of the
duction with modern instruments on thebasis of the results obtained from the testfield photographs.
GENERAL ASPECTS OF HIGH ALTITUDEPHOTOGRAPHY
PHOTOGRAPHIC AIRCRAFT
Recently, very high flying small jet aircraft
ABSTRACT: Cameras equipped with the most recently developedlenses from Wild Heerbrugg Ltd. were used to take aerial photographs from small jet aircraft over the test range at Casa Grande,Arizona. Aircraft, flying height,field angle, and photographic emulsion were varied in the tests. Numerous stereopairs were evaluatedin a precision plotter and under a P.l. microscope. The height accuracy obtained from 15 em (6") focal length lenses was 0.004 percentof the flying height or one part in 25 000, which is a new achievement in analog photogrammetry. Systematic effects varying withaltitude were detected, and a method for compensation was developed. Ground resolution was better than 2.5 m on the ground forthe maximum flying height of 13.4 km (44 000 ft.).
The enormous area coverage, combined with high geometric accuracy and ground resolution of high altitude photography, permit theuse of new procedures in aerotriangulation, and in small-scale lineand ortho-photo mapping. Some proposals for new procedures inthese fields of application are made. Other aspects of high-altitudephotography, such as automatic navigation, semi-automatic operation of a dual camera system, camera performance, and installationrequirements, are discussed as well. The four photographicairplanes tested are now in use in South America.
authors have installed camera systems andparticipated in jet test flights. Numerousphotographs were taken by D. Gut over atest field. The image material was analyzedwith regard to geometric accuracy and imagequality. The purpose of this paper is to showthe unique characteristics ofjet photographyand to determine the potential for map pro-
with pressurized cabins have been used repeatedly by civil organizations and privatecompanies for aerial photography. The Learjet of the Gates Learjet Company (USA), forexample, has a service ceiling up to 13.7 kmalong with other advantages important forphotogrammetric application. The highrate-of-climb performance (in less than 20
PHOTOGRAMMETRIC EGINEERING AND REMOTE SENSING,
Vol. 43, No. 10, October 1977, pp. 1245-1255.
1245
1246 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING, 1977
minutes up to 12 km), the high velocity(about 800 kmlh), and the range of about 3000km rapidly bring the camera system to thetarget area at high flying heights. Large areasmay be flown over at equal illumimition andweather conditions. For photography atlower flying heights the maximum velocitycan be reduced to a third. The flight altitudeis stable and vibrations are slight. Consequently, there are only small tilts and theimage quality is primarily determined by thelens. The working conditions in the pressurized cabin are very comfortable for theoperator. The camera is no longer exposed toexh'eme changes of pressure and temperature. The necessary floor closing glass, however, is palt of the image system. Its quality.also determines the quality of the photograph.
INSTALLATION OF THE CAMERA SYSTEMS
First of all the installation of a window isnecessary in pressurized aircraft. Its dimensions are determined by the field angle ofthe lenses used, the position of the entrancepupil, and especially by the security regulations. In the Gates Lealjet Model 25 C, forexample, 508 x 584 x 38 mm3 windows areinstalled for the RC-10 equipped with asuper-wide-angle lens. The optical quality ofthe windows must be adapted to that of thelenses and filters in order to keep geometricand optical errors small. The two surfacesmust be parallel. Deflection of the glassplate causes distortion. The cameras havedifferent lenses, suspensions, viewfinders,and navigation telescopes. For installation ina jet the most f~1Vorable components have tobe chosen and special measures have to betaken. For the Aviophot System of Wild thefollowing general aspects are of importance:
• The window dimensions are to be fixed forthe lens with the largest image angle. Iflenses with a smaller image angle are alsoused, a spacer has to be used in order toclear the lens cone fi'om the window.
• The rotation of the camera for the purposeof the drift, tip, and tilt correction and itstilting for the change of filters require special distances to the side walls and insidesetups.
• For the exclusive use of the super-wideangle lens a suspension of low height canbe used. The entrance pupil of the lensthus comes as near as possible to the window; its size can then be kept smaller thanwhen using the standard suspension.
• For navigation, drift, and overlap control,the viewfinder telescope can be installedbehind a window. It also can be prepared
for simultaneous control of a second camera (slave camera).
In the Learjet the door of the airplane hasbeen modified for a single camera and canrapidly be exchanged against the standarddoor. 1 For the double camera two windowslying side by side have been installed (compare Figure 1).
Often navigation of a jetplane is no longerperformed with the viewfinder or navigationtelescope, which merely control overlap.The Inertial Navigation System, e.g., theINS-61B of Collins Radio Company or theLTN-51 of Litton Industry, and its couplingwith the autopilot hold the aircraft automatically at a very stable altitude along the desired flight path. Only the coordinates of thewing points of the flying pattern have to bekeyed into the computer of the INS.Moreover, the drift angle of the aircraft isindicated and can be transferred to the camera manually.
PHOTOGRAPHIC PROBLEMS IN HIGH-ALTITUDE
PHOTOGRAPHY
Systematic errors affect photographs takenthrough a window at high altitudes. Theseerrors are named and estimated according totheir size and their consequences.
Earth curvature. Elevations determinedby geodetic surveying relate to the roughlyspherical plane of mean sea level. Consequently, the photogrammetric instrumentalso must have a spherical reference plane orother means of compensation. If no provisions for compensation are incorporated inthe stereoplotter, the photogrammetric models will be warped. Warping depends on theflying height and the focal length of thecamera. The errors are high, as illustrated inFigures 2 and 3.
Atmospheric refraction. The index of refraction of air decreases with increasing al-
FIG. 1. Dual camera installation in the Learjet.
HIGH ALTITUDE PHOTOGRAPHY: ASPECTS AND RESULTS 1247
FIG. 2. Influence of earth curvature on the imageposition of a wide angle photograph if = 153 mm)as a function of the flying height. The linear portion (scale) has been removed in that presentationas distOition.
titude due to changes in temperature andpressure. The imaging ray will be bent andthe image point displaced towards the imageedge. However, the errors are very small andof opposite sign to that of the earth curvature
(Figures 4 and 5).Window flexure and change of refraction
index of air. At high altitudes temperatureand pressure are very different on the twosides of the window. The resultant Hexure ofthe window and difference in the refractionindex of air produce a displacement of theimage. However, the amount of the displacement is very small if the window ismanufactured and installed properly (Figures 6 and 7).
Change in focus and calibrated focallength. In general, lenses with longer focallengths (f=305 mm,f=213 mm) are focusedby the manufacturer to finite object distances, for example, to 850 m at the WildAviotar At II. If the lenses are used at highaltitudes, the resolving power may decreaseunder certain conditions. Especially the 305mm lens is critical when maximum apertureand black-and-white films of high resolvingpower (e.g., Kodak Plus X) are used at altitudes above 4 km. In this case the use ofsmaller apertures or special focusing onlonger object distances is recommended.The above-mentioned image displacementsdue to window Henue and atmospheric reh.·action lead to a change of the calibratedfocal length. These values are smaller than0.02 mm and can be neglected in most cases.
[mmJ
~..~.
• 5km.-'40
20
30
10
- SO
- 20
- 10
- 40
- 30
max. hei hterror
Iml
10
r%0 hI'-----------1+-0--------1+S-f-1,;.,Y..i ng [::;9ht h
flying height h
S 10 lS [kml
FIG. 3. Maximum vertical errors in the middle of the neat model dueto earth curvature as a function of flying height and focal length. Forward lap 50 percent, sidelap 0 percent, control points in the four corners of the model.
1248 PHOTOGRAMMETRIC ENGI EERI G & REMOTE SENSING, 1977
d'
- J
.~l-.,
./0
I •• - 1
- 2
- 3
- 4
Imj
FIG. 4. Displacement of the image due to atmospheric refraction as a function of the image radiusand the flying height. Values are based on awide-angle photograph if = 153 mm).
PRACTICAL TESTS
The functional tests of the Learjet photographic aircraft:, which are carried out on aroutine basis, include taking photographyover a test range. In the last two-and-onehalf years several tests have been performedin which aircraft, camera, lens, altitude, andseason were varied. Numerous stereopairshave been evaluated with respect to geometric accuracy and image quality. From thisexperience the current procedures in mapping at small and medium scales shall beexamined. The conditions during the testsshall first be explained.
PHOTOGRAPHIC AIRCRAITS
The Lealjet models 24 D, 25 B, 25 C, and35 C were used. A single camera was installed in the interchangeable door or a doublecamera in the fuselage. In the latter case asliding door protected the windows duringtake ofT and landing. The window was a 51cm x 58 cm x 3.8 em Schott BK 7 glass plate.The pressure inside the cabin correspondedto the pressure at an altitude of 2 km.
TEST RANGE
The 25 km x 25 km Casa Grande test
FIG. 6. Displacement of the image due to window flexure and difference of the refraction indexof air at both sides of the window. (Calculationwith the following parameters according to reference 2: Index of refraction of glass n = 1.517,temperature inside cabin T = 20°C, pressure inside cabin corresponding to an altitude of 2 km,window dimension 58 x 51 x 3.8 cm3 , ElasticityModul of glass 0.7' 106 kp/cm2 , focal length oflensf = 153 mm, position of entrance pupil of the lensabove window ISO mm)
range near Phoenix, Arizona was used. Itsdensity of control points, one target per 2.6km2 , and the shape of the panels allow measurements with photography of varying andextremely small scales. The target (see Figure 8) has four quadrilateral "pointers," theparallel sides of which measure between 0.8m and 4.9 m. The smallest visible image distance in the photograph can be determinedapproximately with the help of this panel bya simple procedure (see Figure 15).
Camera system. For photography, theWild Heerbrugg Aviophot System was used.The main part of the system is the RC 10camera, in which six lens cones can be introduced. In the tests the following lens coneswere used:
• the super-wide-angle lens 8.8SAG II, f =88.5 mm, Amax = 5.6
• the wide-angle lens I5UAG I,f = 153 mm,Amax = 5.6
/lt1lllall 1m
] f .. 153 llIIl
I'" hi0.6 ~/.
~ 0.05 0.5 /~
0.4 ~:r-'- 0.03 O.~.........--....................
0., ,//
• 0.01 0.1
.lhm.tx
I'" h] Iml( .. 153m. -'+ 0.03 o.3~:>, < - .
.0.02 0.2 _---.--
• 0.01 0.1--
FIG. 5. Maximum vertical error within the neatmodel of wide-angle photographs if = 153 mm)due to atmospheric refraction.
1513
10flyin9 height h
lkol
FIG. 7. Maximum vertical error within the neatmodel of wide-angle photographs if = 153 mm)due to window flexure and change in the refractive index of air.
flying height h
(""I1513
10
HIGH ALTITUDE PHOTOGRAPHY: ASPECTS AND RESULTS 1249
5 m
FIG.8. Target of the Casa Grande Test Range. Itssize and shape are suitable for photography ofhigh-altitude and of vmying photo scales.
• the wide-angle lens 15UAG II,! = 153 mm,Am.x = 4
• the normal angle lens 21NAG II, f = 213mm, Am.x = 4
The last three lenses have been put on themarket just recently. The lenses have a verysmall distOition which causes a correspondingly small deformation of the stereomodel(see, for example, Figure 9). The singlecameras were controlled either by a navigation sight or a viewfinder in a separatemount. The double camera used is shown inFigure 10.
Both of the cameras are controlled by onenavigation sight; they work simultaneouslyor independently with different lenses, over-
lap, and film. Using a dark yellow filter, twotypes of films, the panchromatic film KodakPlus X and the false color film Kodak Aerochrome Infrared, were exposed. Exposurewas automatically controlled.
PHOTOGRAPHIC FLIGHTS
Test flights were carried out over the CasaGrande test range with four Learjets for general installation and camera tests and inorder to determine the geometric accuracy.Photographs of the test range were takenfrom altitudes of 5 km, 8.9 km, 11.9 km, and13.4 km on four different dates. Because thelens cones were exchanged during the tests,image scales between 1:33 000 and 1: 150000 were obtained. Navigation was carriedout with the Inertial Navigation System,with the help of which the drift angle wasobtained and manually transferred to thecamera-controlling navigation sight. Thismethod proved to be very practical and accurate. The overlap, however, was controlledin the usual way. For manual navigation theforward view of the navigation sight did nothave to be used at these altitudes. Therefore,the camera's viewfinder, installed in a separate mount, may be used as well. Its installation and use behind a window pose no problems.
- 1
- 2
[mml
MEASUREMENT OF PHOTOGRAPHY
The measurement of single models wascarried out on a precision stereocomparatorand on two precision stereoplotters, the WildSTK 1 and the Wild A 10, respectively. Theaverage elevation accuracy of the two analoginstruments used in the measurements is mil
nIll
30 60 90 r:rn
FIG. 9. Radial distortion of the wide-angle lensUAg II and the resulting height deformation (values in micrometers and related to the photograph,overlap 61 percent). 3 /Lm = 0.02 °/00 h.
FIG. 10. Lemjet dual camera installation. Fromleft to right: master camera with 15 cm lens andspacer, navigation telescope with overlap regulator, slave camera with 8.8 cm lens, circuitryunit, control unit, and exposure meter.
-3 -2 -2 -3
-3 0 0 -3
-2 1 -1 -2
0 2 2 -i0 0 0 0
0 0 0 1
0 0 0 0
0 2 2 -1
-2 1 1 -2
-3 0 0 -3
-3 -2 -2 -3
100
80
60
40
20
o20
40
60
80
100
RESULTS
GEOMETRIC ACCURACY
In Table 1 the measuring results of tensingle models obtained by the precisionstereoplotter are shown. They comprise theerrors of planimetry and elevation on theground and in the image.
Let us first discuss the results with theeight wide-angle stereopairs. The averageRMSE in elevation of ± 0.034 %0 of the flyingheight, or 1:33 000, is remarkably small. Thisis almost the limit of performance for measurement with an analog instrument. Themean errors of planimetry average mX,y= ± 10 /-Lm in the photograph. Consequently,the accuracy of planimetry is inferior to theaccuracy of elevation. The reason for this maybe the partially bad quality of the signals andthe resulting insecurity in setting themeasuring mark. Elevation setting could always b~ done in the horizontal neighborhood of the target. The influence of the flying height on the vertical error is shown inFigure 11.
The accuracy in elevation is best at thelowest flying height (mil = ± 0.026 0/00 ofthe flying height). At flying heights of 9 kmand more the accuracy of elevation is aboutthe same (0.038 °/00 of h). The elevation accuracy of the measured normal angle andsuper-wide-angle stereopairs is inferior tothat of the wide-angle stereopairs; itamounts to an average of mIt = ± 0.07 °/00 ofthe flying height. Better results for thesuper-wide-angle photographs can be obtained if compensation plates for lens distortion are used. When measuring thestereopairs, systematic model deformationsl!-P to about 0.2 %0 of the flying height werefound. This error, however, could largely becorrected by increasing the model scalenumber to be set in the compensation devicef:Or the ealth curvature. The magnificationfactors K, empirically found for the wideangle stereopairs, are shown in Figure 12; alinear dependence on the flying" heightexists. Some stereopairs also were evaluatedanalytically in order to detect the reason forthe systematic errors. The model deformation was found to be of the same size (see inFigure 13). This fact proves the good adjustment for the stereoplotter used and for itscompensation device for earth curvature.The reason for the deformation is very likelya higher flexure of the window during theflight than theoretically expected.
* Parameters of the instrument test are measurement of 66 grid points, f = 150 mm, Z
= 450 mm, bx = 300 mm, and MM = 1:100 000.
1250 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING, 1977
= ± 0.021 %0 of the, flying height.* Thephotographs are observed at 8 x magnification. The instrument features a compensation device for the influence of earth curvature. A sphere built in the model space iscontinuously scanned so that the referencein height will change with the position in themodel. The sensed value will be exaggerated by a factor corresponding to the flyingheight or the model scale denominator.
If systematic errors of a spherical natureremain, the compensating efJect can be increased or decreased by changing the modelscale denominator. In this way systematicerrors in the stereo model, e.g., those due toatmospheric refraction, lens distortion, andflexure of the camera's pressure platen or theaircraft window, may be approximatelycompensated.
The revised model scale denominatorwhich is set in the correction device for eartl~curvature, has to be determined empirically.
The stereoplotter was equipped with adata acquisition device, the Wild EK 22,with connected desk calculator HP 9820.The coordinates and the elevations of eachpoint were measured three times. The averaged ,model coordinates were immediatelytransformed into ground coordinates at theside of the analog instrument. A simple spatial transformation over four conb'ol pointseliminates the small errors frequently arising in manual absolute orientation. Thenumber and arrangement of control pointsvaried.
The stereocomparator STK 1 is an evenmore accurate measuring instrument (mX,y= ± 1 /-Lm). The measurement is done in amodel affected with parallaxes. The continuous removal of vertical parallaxes at eachpoint calls for a high quality of the imagedtargets if the high accuracy of the comparatoris to be exploited. Systematic image errorssuch as lens distortion, earth curvature, rehaction, etc., are compensated for by calculation. Determination of the coordinates ofthe test field was done with the help of acomputer.
The photointerpretation stereomicroscopeWild M 5 with parallel guided and illuminated picture carriages was used for a comparing examination of the image quality. Thechosen enlargement of viewing was 12.5 x.The ocular was equipped with a rotatablemeasuring scale.
HIGH ALTITUDE PHOTOGRAPHY: ASPECTS AND RESULTS 1251
TABLE I. ACCURACY Of HIGH-ALTITUDE PHOTOGRAPHY
Aircraft Date Photo Scale RMSE2Control
Lens Cone Flying Height' Photo No. Pts. Terrain3 Photograph4
25C Sept. 74 1:33000 8 mE = ± 0.24 m = + 7x -15 UAgII 5000 ill 1345/46 fiN = ± 0.39 m" = ± 12No. 3035 (16400') mH = ± 0.13 m, = ± 0.026
25C Sept. 74 1:33000 8 mE = ± 0.23 m = + 7x -15 UAgII 5000 ill 1341/42 mN = ± 0.37 m,,=±11No. 3035 (16400') fiH = ± 0.13 1Ylz = ± 0.026
25B Aug. 75 1:56 500 10 mE = ± 0.75 mx = ± 138.8 SAgII 5000 ill 180/81 mN = ± 0.54 m y =±10
o. 2073 (16400') mH = ± 0.36 m, =± 0.07225C Sept. 74 1:58000 28 mH = ± 0.37 m, =± 0.042
15 UAgII 8800 ill 1356/57 ± 0.475 ± 0.0535
No. 3035 (29000')24D jan. 75 1:57000 26 fiH = ± 0.30 m, =± 0.036
15 UAgII 8400 ill 79/80No. 3047 (27750')
25B Aug. 75 1:58 000 28 fiH = ± 0.33 m, =± 0.03715 UAgI 8900 ill 118/19 ± 0.665 ± 0.0745
No. 6034 (29000')25B Aug. 75 1:58 000 28 mE = ± 0.66 mx =±l1
15 UAgl 8900 ill 119/20 mN = ± 0.70 m" = ± 12(29000') 1Ylz = ± 0.034
25C Sept. 74 1:88000 11 mE = ± 0.54 11~x = ± 615 UAgII 13400 ill 1273n4 mN = ± 0.92 m,,=±11No. 3035 (44 000') mH = ± 0.52 1Ylz = ± 0.039
25C Sept. 74 1:63 000 10 mE = ± 0.54 m = + 9x -21 NAgII 13400 ill 56/57 fiN = ± 0.82 m" = ± 13
(44000') fiH = ± 0.78 m, =± 0.06035C May 76 1:78000 44 fiH = ± 0.41 1Ylz = ± 0.035
15 UAgII 11 900 ill 68/69No. 3078 (39000')
1 Above mean ground level2 Root-mean-square error3 m1;. m;v, mJl in meters4 m,e. mll in the micrometers, m.= in °/00 of the flying height.S analytical evaluation
0.10 ~oo H
21NAG II 0
~15UAG 1/11 0
0 8.8SAG II •" •c0.~
~
0rn>~
.05 ~
~
c~ 0-:,--------- - ---0 0--------0-~ 0 0co
00
flying height
10 15 k..
FIG. 11. Elevation errors as a function of the flying height and the lens.
1252 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSI G, 1977
15UAG II 6034 • 3047
2.03078 @ 3035 o
10
1.5
1.0flying height
--0 ------ ... __ 0-- ... 0
---- ------""'@>--"'-.o,
15 km
modelscale 1: 15 000 25 000 40 000
FIG. 12. Systematic errors in high-altitude photography can be compensatedby using an increased altitude or model scale number used in earth curvaturecorrection. The factor of enlargement K varies with the altitude. The circles inthe diagram indicate empirically derived values for four wide-angle lenses infour different Learjet photographic aircraft.
IMAGE QUALITY
The procedure and the results of theimage quality evaluation are shown in Figures 15 and 16. The smallest visible imagedistance is influenced by many factors, e.g.,the contrast of the target to its neighborhood,the location in the image, the type of lensand the photographic emulsion, and others.Five different panels named S, R, D, H, andW were measured on 14 photographs at
scales between 1:56000 and 1:99 000. Theirlocation in the image was up to 130 mm tromthe center. This fact, however, influencedthe results only slightly.
Considering the measurements first, onecan see that in 75 percent of the cases thesmallest imaged distances are less than 30/-lom, that is, 1.8 m (6 ft) at the 1: 60000 photoscale. At the extremely small image scale of1:150 000 the panels with good contrast of
Deviation in2 m elevation Profile AG
"".... .........., ........ ....
'",/
'" --- 10 km
'"",
50 49 48 47 46 45 44
2 Deviation inm elevation Profile AH
................ _----"
....10' .... km
"50 49 48 47 46 45 44
FIG. 13. Example of the defom1ation of a high-altitude stereo model.
Data: Photographic aircraft LeaIjet 25-CFlying altitude 8900 mCamera RC 10 with
UAg II No. 3035Stereopair 1356/57Overlap 55 percentMeasurement in A 10 (dashed line)
and STK 1 (solidline) by D. Gut
Revised altitude forcorrecting earthcurvature will be 13200 m
The positions of the profiles are shown in Figure 14.
HIGH ALTITUDE PHOTOGRAPHY: ASPECTS AND RESULTS 1253
AF AG AH AJ
46
45
44
43
FIG. 14. Residual errors in elevation of a high-altitude stereo model.Data in Figure 13. Measurement in the A 10 stereoplotter
the Casa Grande test range (width b = 4.9 m)were still visible. Comparing the lens types,no diflerence can be detected for the threelenses. False color HIm shows equally goodresults as the panchromatic film. However, agreater percentage of targets were visible onthe false color film. Obviously, the lower resolution of false color film is compensated bythe higher contrast of this HIm.
Inflight camera tilt. Camera tilt at themoment of exposure was determined duringmeasurement in the stereoplotter. For sevenstereomodels the maximum range of tilt wasbetween
-0.55" to +1.65" for lateral tilt (w)-0.30" to + 1.06" for longitudinal tilt (</J)
These values are relatively small. This factproves the stable flying altitude of jet aircraft. Small camera tilts will favorably influence the measurements in analog stereoplotters.
ApPLICATION OF HIGH-ALTITUDE
PHOTOGRAPHY IN TOPOGRAPHIC MAPPING
The enormous area coverage, combinedwith high geometric accuracy and ground resolution of high-altitude photography, permits the use of new procedures in aerotriangulation, and in small-scale line and orthophoto mapping. As early as 1969 the potential of high-altitude aircraft for topographicmapping was impressively documented byR. E. Altenhofen with the help of practicalexamples. 3 In the meantime, new instrumentation for picture-taking and measurements became available, e.g., handwheel and freehand driven precisionstereoplotters and digitally controlled orthophoto printers of high resolution and highenlargement capability. With the help of thisnew instrumentation and the experiencegained from the evaluation of the photog-
Tar~et
f-0.81~
Photo~raph
Ib = 0.01 + 4.07 (7.62 - Y) values in m
7.62
FIG. 15. Detennination of the smallest visible image distance bymeans of a quadrilateral target. The length y of the imaged targetwill be measured, the width b will be detennined by calculation.Values in m.
1254 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING, 1977
• Black & White rillllo false Color film
II'"I Iml'\ '" 1 :60 000
45
2.540
35 002.0
0
30
25 1.5•• 0
O'
20
1.015 ....10
0.5
tanJel S , o H W
lens UAG I UAG II SAG II
FIG. 16. Smallest visible image distance determined for three types of lenses and two photographic emulsions with the help of five differentpanels on 14 photographs of scales between 1:56000 and 1:99000. The panels measured were up to130 mm hom photograph center. The values arethe mean fi'om the four target pointers.
raphy over the Casa Grande test range, thecurrent procedures in small-scale mappingwill again require discussion.
AEROTRIANGULATION
With maximum Hying height and superwide-angle lenses, the distance betweenfield control points may be increased. Whenusing wide-angle photographs of 1:80 000scale, precision stereoplotters or comparators, and modern computationalmethods, a distance of about 60 km can bebridged. 4 The errors in planimetry and elevation of the control points should be smaller than 2 m. Maintaining this accuracy, thebridging distance can be extended to about100 km when using super-wide-anglephotographs taken from maxim um Hyingheight. Similarly, the number of photographs is considerably reduced. An area of100 x 100 km2 , e.g., may be covered by only36 super-wide-angle photographs.
MEDIUM AND SMALL SCALE LINE MAPS
Line maps at scales between 1:50000 and1:25 000 had until recently been mappedfrom photography at scales which were small-
er than or the same as the map scale. Thelimitations in photo scale were conditionedby the service ceiling of small aircraft: and thepoor working climate in unpressurized cabins. When using small jet aircraft, the scalelimitations are determined by the adequacyof interpretation of small details and by thecontour interval to be mapped. But with thehelp of precisionplotters like the A 10 or thenew generation of analog stereoplotters suchas the Aviomap AMH, in which
• the measuring mark is conh'olled by handwheels and fi'eehand,
• the photograph is observed at high andchangeable magnification, and
• the plotting is done with the help of an au-tomatic table at the operator's side,
photography of smaller scales than beforecan be interpreted and accurately be measured. When using the new generation ofwide-angle cameras and modern precisionstereoplotters, the photo scales listed inTable 2 seem to be applicable for open terrain.
When using modern super-wide-anglecameras, both map scales may be producedb'om 1: 100 '000 photography (b = 3 m, t:.h = 5m).
For planning purposes with low-accuracyrequirements, maps of larger scales such as1:5 000 could be mapped efficiently homhigh-altitude photography as well.
Orthophoto mapping. Orthophoto mapsare produced in large quantities at scales of1:25 000 (1:24 000) and 1:10 000. Theireconomic production is determined by,among other things, the ratio photo-to-mapscale. Modern orthophoto printers and goodphotography allow enlargement factors between 4 X and 6 x. Limitations for the factorarise £i'om the service ceiling of the photographic aircraft: and the smallest detail to beprinted and recognized by the user. In addition to the wide-angle camera, the normalangle camera with f = 21 cm is preferred insteep and wooded areas. For the map scalesconsidered, such a camera leads to altitudeswhich can be reached by jet aircraft only. Inopen and hilly terrain, super-wide-angle
TABLE 2. PROPOSED PHOTOSCALES FOR MEDIUM- AND SMALL-SCALE LINE MAPPING. THESE SCALES
WILL BE ACHIEVED FROM ALTITUDES OF 9 AND 12 km ABOVE TERRAIN, RESPECTIVELY. IN THE
LATTER CASE THE TERRAIN MAY STILL, BE Up TO 1.5 km ABOVE SEA LEVEL.
Map Scale
1:250001:50 000
Photo Scale
1:60000 - 8000080000
Smallest Visible ImageDistance b
2.0 - 2.5 m2.5m
Smallest ContourInterval M
4-5m5m
HIGH ALTITUDE PHOTOGRAPHY: ASPECTS AND RESULTS 1255
TABLE 3. PROPOSED PHOTO SCALES FOR MEDlUM- AND SMALL-SCALE ORTHOPHOTO MAPPING.
1:10 0001:25000
f= 88.5 mm
30 000 - 50 000100 000
f= 153 mm
40 000 - 60 00080000
f= 213 mm
6000060000
photographs can be utilized provided thatthe orthophoto printer accepts such photographs and compensates for slope across thedirection of scanning. For the photomapsconsidered, the photo scales at optimal fieldangle listed in Table 3 seem to be practicable.
The camera should be selected accordingto the terrain. The scales were based on theassumption that the smallest detail to be recognized in the photomap has a width of 2 mfor 1: 10 000 and 3 m for 1:25 000. Theplanimetric accuracy required for the mapswill easily be met by modern orthophotoprinters.
ACKNOWLEDGMENT
The authors are obligated to several persons for their support in this investigation.Dipl.-Ing. S. Heggli, Institute for Geodesyand Photogrammetry, E. T. H. Ziirich, performed the analytical point determinationfrom the stereocomparator data. Dipl.-Ing.Kreutzer, Wild Heerbrugg, calculated themodel deformations due to distortion, refrac-
tion, and window. The Optical CalculationDepartment of Wild Heerbrugg supportedthe authors in determining the image errorsdue to the bent window. Furthermore, Mr.Dipl.-Ing. Z. Pm'sic, E. T. H. Ziirich, shouldbe thanked for his help in comparator measurements; Mr. S. Blankenship of the GatesLealjet Company for the organization of thephotographic flights; and the Defense Mapping Agency, Washington, D. c., for providing the coordinates of the Casa Grande testrange.
REFERENCES
1. Bock, A. G., A breakthrough in high altitudephotographs from jet aircraft, Photogrammetric Engineering, Vo!. XXXIV, No.7, July1968.
2. Hiitte, Des Ingenieurs Taschenbuch, 28.AutI., S. 946, 1955.
3. Altenhofen, R. E., Application of high altitudephotographs in topographic mapping, PanAmerican Institute of Geography and History, Washington, D. c., 1969
4. US Geological Survey, Annual report ofUSGS Cartographic Research, June 1976.
Articles for Next MonthU. V. Helava, The Analytical Plotter-Its Future.Gottfried Konecny, Software Aspects of Analytical Plotters.Dr. Bernard L. Y. Dubuisson, Why Analytical Plotters?Dr.-Ing. D. Hobbie, C-lOO PLANICOMP, the Analytical Stereoplotting System from
Carl Zeiss.Christian Vigneron, TRASTER 77: Matra Analytical Stereoplotter.Richard H. Seymour, The US-l Universal Stereoplotter.R. E. Kelly, P. R. H. McConnell, and S.]. Mildenberger, The Gestalt Photomapping System.S. Jack Friedman, The OMI Analytical Stereoplotter Model AP/C4.Prof. Giuseppe lnghilleri, The D.S.-Type Calileo Analytical Plotters.