Principles of Photogrammetry
Principles of Photogrammetry
Principles of Photogrammetry
31 Mar 93
10-2. General
10-3. Definition
10-1
EM 1110-1-1000
31 Mar 93
where
S = photographic scale at a point
10-2
EM 1110-1-1000
31 Mar 93
The flat terrain assumption, however, introduces scale This solution is not an approximation because the effect
variation errors. For accurate determinations of horizon- of scale variation caused by unequal elevations is in-
tal distances and angles, the scale variation caused by cluded in the computation of the ground coordinates. It
elevation differences between points must be accounted is important to note, however, that the elevations ha and
for in the photogrammetric solution. hb must be known before the horizontal ground coordi-
nates can be computed. The need to know elevation can
b. Horizontal ground coordinates. Horizontal be overcome if a stereo solution is used.
ground distances and angles can be computed using
coordinate geometry if the horizontal coordinates of the c. Relief displacement. Relief displacement is an-
ground points are known. Figure 10-2 illustrates the other characteristic of the perspective geometry recorded
photogrammetric solution to determine horizontal by an aerial photograph. The displacement of an image
ground coordinates. point caused by changes in ground elevation is closely
related to photographic scale variation. Relief displace-
(1) Horizontal ground coordinates can be calculated ment is evaluated when analyzing or planning mosaic or
by dividing each photocoordinate by the true photo- orthophoto projects. Relief displacement is also a tool
graphic scale at the image point. In equation form, the that can be used in photo interpretation to determine
horizontal ground coordinates of any point are given by heights of vertical objects.
xp(H hp) (1) The displacement of photographic images caused
Xp
f by differences in elevation is illustrated in Figure 10-3.
(10-4) The image displacement is always along radial lines
yp(H hp) from the principal point of a truly vertical photograph or
Yp
f the nadir of a tilted photograph. The magnitude of
relief displacement is given by the formula
where rh (10-6)
d
H
Xp,Yp = ground coordinates of point p
Note that these equations use a coordinate system de- r = radial distance from the principal point to the
fined by the photocoordinate axes having an origin at image point
the photo principal point and the x-axis typically
through the midside fiducial in the direction of flight. H = flying height above ground
Then the local ground coordinate axes are placed paral-
lel to the photocoordinate axes with an origin at the (2) Since the image displacement of a vertical object
ground principal point. can be measured on the photograph, Equation 10-6 can
be solved for the height of the object to obtain
(2) The equations for horizontal ground coordinates
are exact for truly vertical photographs and typically d(H hbase)
ht (10-7)
used for near vertical photographs. rtop
10-3
EM 1110-1-1000
31 Mar 93
10-6. Exterior Orientation of Tilted Photographs Figure 10-4. Exterior orientation of an aerial
photograph
Unavoidable aircraft tilts cause aerial photographs to be
exposed with the camera axis tilted slightly from verti- b. The tilt-swing-azimuth system is appropriate for
cal, and the resulting pictures are called tilted photo- hand calculations. An auxiliary photocoordinate system
graphs. The equations given above are exact for truly is defined with an origin at the photo nadir and y axis
vertical photographs, and they are used with near verti- along the direction of tilt. The expression for scale on a
cal photography for planning, estimating, and photo tilted photograph is
interpretation. However, an accurate photogrammetric
solution using aerial photographs must account for the
camera position and tilt at the instant of exposure. f
(y sin t)
cos t (10-8)
S
a. The exterior orientation of a photograph is its H h
spatial position and angular orientation with respect to
10-4
EM 1110-1-1000
31 Mar 93
10-5
EM 1110-1-1000
31 Mar 93
10-8. Geometry of Aerial Stereophotographs where d is the negative format dimension. The flying
height above datum is also found using average scale
The basic unit of photogrammetric mapping is the stere- and average ground elevation.
omodel formed in the overlapping ground coverage of
successive photographs along a flight line. Figure 10-5
illustrates the ground coverage along three contiguous f
H have (10-10)
flight lines in a block of aerial photographs. Along each Save
flight line, the overlap of photographs, termed end lap,
is typically designed to be 60 percent. End lap must be
at least 55 percent to ensure continuous stereoscopic (2) Let B represent the air base between exposures
coverage and provide a minimum triple overlap area in the strip. Then from the required photo end lap Elap
where stereomodels can be matched together. Between
adjacent flight lines, the overlap of strips, termed side
Elap (10-11)
lap, is typically designed to be 30 percent. Side lap B G 1
must be at least 20 percent to ensure continuous stereo- 100
scopic coverage.
Slap (10-12)
W G 1
100
10-6
EM 1110-1-1000
31 Mar 93
where
p = parallax
10-7
EM 1110-1-1000
31 Mar 93
10-8
EM 1110-1-1000
31 Mar 93
c. Instrumentation
The collinearity condition equations include all interior (2) Stereoplotters. Some specialized stereoplotters
and exterior orientation parameters required to solve the have been developed specifically for terrestrial applica-
resection and intersection problems accurately. Analyti- tions; however, these instruments are not typically
cal solutions consist of systems of collinearity equations available in the United States photogrammetric industry.
relating measured image photocoordinates to known and Some analog stereoplotters can accommodate some
unknown parameters or the photogrammetric problem. terrestrial and close-range cameras; but often incompati-
The equations are solved simultaneously to determine ble focal lengths, format sizes, and depth of field con-
the unknown parameters. However, since there are straints make analog stereoplotters unusable for this
usually redundant measurements producing more equa- work. Analytical stereoplotters can accommodate terres-
tions than there are unknowns in the problem, a least trial and close-range cameras because no mechanical
squares adjustment is used to estimate the unknown constraints are placed on the mathematical projection.
parameters. The least squares adjustment algorithm Of course, monocomparators and analytical software
includes residuals vx and vy on the measured photocoor- work equally well for precise applications of point
dinates that estimate random measurement error. mapping.
10-9
EM 1110-1-1000
31 Mar 93
10-10
EM 1110-1-1000
31 Mar 93
accurate and precise image measurement. The x and y These instruments will display a digital image on a
photocoordinates are encoded, and all interior and workstation screen in place of a film or glass diaposi-
exterior orientation parameters are included in the tive. The instrument will operate as an analytical
mathematical projection model. Except for the positive stereoplotter except that the digital image will be viewed
format size that will fit on the photo stage, the analytical and measured. The accuracy of digital stereoplotters is
stereoplotter has no physical constraints on the camera governed by the pixel size of the digital image. The
focal length or model scale that can be accommodated. pixel size directly influences the resolution of the photo-
coordinate measurement. A digital stereoplotter can be
(1) The viewing system is an optical train system classified according to the photocoordinate observation
typically equipped with zoom optics. The measuring error at image scale. Then it should be comparable to
mark included in the viewing system may be changeable an analog or analytical stereoplotter having the same
in style, size, and color. The illumination system should observation error, and the standards and guidelines in
have an adjustable intensity for each eye. this manual should be equally as applicable. See also
paragraph 10-15.
(2) The measuring system consists of an input
device for the operator to move the model point in three 10-14. Stereoplotter Operations
dimensions. The input device is encoded, and the digi-
tal measure of the model point movement is sent to the All stereoplotters, whether analog or analytical, must
computer. The software then drives the stages to the be set up for measuring or mapping in three
proper location accounting for interior and exterior consecutive steps: interior orientation, relative orien-
orientation parameters. These operations occur in real tation, and absolute orientation.
time so that the operator, looking in the eyepieces, sees
the fused image of the floating mark moving in three a. Interior orientation. Interior orientation in-
dimensions relative to the stereomodel surface. The volves placing the photographs in proper relation to
operators input device for model position may be a the perspective center of the stereoplotter by match-
hand-driven free-moving digitizer cursor on instruments ing the fiducial marks to corresponding marks on the
designed primarily for compilation, or it may be a hand photography holders and by setting the principal
wheel/foot disk control or similar device on instruments
distances of the stereoplotter to correspond to the
supporting fine pointing for aerotriangulation.
focal length of the camera (adjusted for overall film
shrinkage).
(3) Analytical stereoplotters are more accurate than
analogic stereoplotters because the interior orientation
parameters of the camera are included in the projection b. Relative orientation. Relative orientation in-
software. Therefore, any systematic error in the photo- volves reproducing in the stereoplotter the relative
graphy can be corrected in the photocoordinates before angular relationship that existed between the camera
the photogrammetric projection is performed. Correct- orientations in space when the photographs were
ing for differential film deformations, lens distortions, taken. This is an iterative process and should result
and atmospheric refraction justifies measuring the photo- in a stereoscopic model easily viewed, in every part,
coordinates to accuracies of 0.003 mm and smaller in without "y parallax"the separation of the two
analytical stereoplotters. To achieve this accuracy, the images so they do not fuse into a stereoscopic model.
analytical stereoplotter must have the capability to per- When this step is complete, there exists in the stereo-
form a stage calibration using measurements of refer- plotter a stereoscopic model for which three-
ence grid lines etched on the photo stage. dimensional coordinates may be measured at any
point; but it may not be exactly the desired scale, and
(4) Analytical stereoplotters capable of precise map it may not be levelwater surfaces may be tipped.
compilation and aerotriangulation measurements include
(but are not limited to) the Zeiss Planicomp P series, the c. Absolute orientation. Absolute orientation uses
Wild (now Leica) AC and BC series, the Kern (now the known ground coordinates of points identifiable
Leica) DSR series, the Intergraph Intermap, and the in the stereoscopic model to scale and to level the
Galileo Digicart (Figure 7-5). model. When this step is completed, the X, Y, and Z
ground coordinates of any point on the stereoscopic
d. Digital stereoplotters. The next generation of model may be measured and/or mapped.
stereoplotters will be digital (or soft copy) stereoplotters.
10-11
EM 1110-1-1000
31 Mar 93
10-12
EM 1110-1-1000
31 Mar 93
is the angle measured in the plane of the equator from local east, north, and up (along the normal), respec-
the reference meridian through Greenwich, England, to tively, at the point. Local rectangular coordinates will
the meridian through the point. Latitude is the angle in appear more natural with respect to the terrain, and the
the plane of the meridian through the point from the coordinate magnitudes will be smaller for computations.
equator to the normal through the point.) A geodetic Photogrammetric aerotriangulation should be computed
height can also be defined as the distance along the in a local rectangular system (especially for large project
normal from the surface of the ellipsoid to the point. areas) since earth curvature and map projection distor-
tion are not present in Cartesian coordinates.
(2) Geocentric coordinates. Geocentric coordinates
are defined for a given datum by placing the origin of a (4) North American Datum of 1927. In the United
right-handed spatial rectangular coordinate system at the States, two horizontal datums are commonly
center of the reference ellipsoid. The z-axis coincides encountered in engineering surveying and mapping: the
with the semiminor axis of the ellipsoid. The x- and y- North American Datum of 1927 (NAD 27) and the
axes are in the plane of the equator with the x-axis North American Datum of 1983 (NAD 83) (Table 10-3).
through 0-degree longitude. Simple Cartesian coordi- NAD 27, a datum defined by an adjustment begun in
nate computations can be performed in the geocentric 1927 of the national control point network, is a conti-
system. The ellipsoid parameters and curved reference nental datum referenced to the Clarke 1866 ellipsoid.
surface are eliminated from the computations The ellipsoid is positioned and oriented to closely fit the
(Figure 10-7). geoid throughout the North American continent. Since
the NAD 27 adjustment, new control surveys were
added to the network by holding the original framework
fixed and adjusting the new survey to fit the original.
As the network grew without readjustment, distortions
were found that demanded a readjustment of the net-
work to meet modern surveying requirements.
10-13
EM 1110-1-1000
31 Mar 93
Table 10-3
Comparison of NAD 27 and NAD 83 Datum Elements
Adjustment Terrestrial data for 25,000 stations and Terrestrial and satellite data for 250,000+
piecewise fits stations
Note:
1. a = length of semimajor axis of ellipsoid.
b = length of semiminor axis of ellipsoid.
f = flattening of ellipsoid.
National Geodetic Vertical Datum of 1929. The eleva- correctly, all areas would retain their correct shape and
tion of a point can be combined with the horizontal relative size, meridians and parallels would intersect at
position on the ellipsoid to define a three-dimensional right angles, and great circles and rhumb lines would be
coordinate, latitude, longitude, and height (, , h), by represented as straight lines. Obviously, the earths
the relationship surface, being spherical, cannot be developed upon a
plane without distortion of some of these quantities.
h H N (10-17)
(2) In engineering surveying and mapping, correct
depiction of shapes is important. Thus, the charac-
teristic of conformality is enforced in projections used
where for large-scale mapping. This is accomplished by math-
ematically constraining the projection scale factor at a
h = height above the ellipsoid point, whatever it may be, such that it is the same in all
directions from that point. This characteristic of a pro-
H = elevation above the geoid jection preserves angles between infinitesimal lines.
That is, all lines on the grid cut each other at the same
N = geoid separation above the ellipsoid angles as do the corresponding lines on the ellipsoid for
very short lines. Hence, for a small area, there is no
c. Map projections. A map projection is a mathe- local distortion of shape. But since the scale must
matical function that defines a relationship of coordi- change from point to point, distortion of shape can exist
nates between the ellipsoid and a plane. The purpose of over large areas. Remember that the distortion present
a map projection is to display the curved earth surface in any map projection is not an "error" since it is a
on a flat plane that can be printed on a sheet of paper. mathematically defined function. The distortion must
If field survey observations are reduced to the plane, the simply be accounted for in transforming data from map
computation and adjustment of horizontal control posi- to ground.
tions can be carried out in two-dimensional Cartesian
coordinates. The need for more involved geodetic sur- (3) The most commonly encountered map projec-
vey computations can be eliminated. tions in engineering surveying and mapping are the
State Plane Coordinate Systems (SPCS). NGS has
(1) The final map reference system is a plane developed SPCSs for each state, and many states have
developed from a regular mathematical surface. Devel- formally adopted the systems by legislative action. The
opable surfaces that can be rolled out into a plane in- SPCSs are based on the Lambert Conformal Conic
clude a plane, a cone, and a cylinder. The ideal map Projection or the transverse Mercator projection. For
projection would display all distances and directions complete discussion of the SPCSs refer to Stem (1989).
10-14
EM 1110-1-1000
31 Mar 93
(4) State Plane Coordinate Systems are defined for b. Purpose. The purpose of aerotriangulation is to
both the NAD 27 and NAD 83 datums. For the extend horizontal and vertical control from relatively
NAD 27 SPCS definition, the unit of length is the few ground survey control points to each unknown
US Survey Foot. For the NAD 83 SPCS definition, the ground point included in the solution. The supplemental
unit of length is variable among the states. Care must control points are called pass points, and they are used
be exercised when using NAD 83 SPCS values in feet to control subsequent photogrammetric mapping.
since either the US Survey Foot or the International
Foot may be used in a specific state or locality. The c. Geometric principles. The aerotriangulation
conversion factors are given in paragraph 2-10a. geometry along a strip of photography is illustrated in
Figure 10-8. Photogrammetric control extension re-
(5) Another commonly used plane coordinate pro- quires that a group of photographs be oriented with
jection system is the UTM projection. Since this is the respect to one another in a continuous strip or block
standard military reference system, operational support configuration. The exterior orientation of any photo-
mapping work may need to be compiled using this graph that does not contain ground control is determined
system. The UTM system uses a Mercator cylindrical entirely by the orientation of the adjacent photographs.
projection with the axis of the cylinder oriented In a given aerotriangulation configuration, each
perpendicular to the earths polar axis. The UTM coor- photograph contributes to the exterior orientation of the
dinate zones are uniformly spaced around the equator at adjacent photographs through the pass points located in
6-degree intervals and extend north and south from the the triple overlap areas. The term triple overlap refers
equator. A useful reference for the UTM system and to the ground area shared by two adjacent stereomodels
other map projections is Snyder (1982). along the strip. The triple overlap area is imaged on
three consecutive photographs. Thus, when end lap is
Section IV specified to be 60 percent, the alternate photographs will
Principles of Aerotriangulation overlap each other by 20 percent at average terrain
elevation.
10-19. General
d. Pass points. A pass point is an image point that
The principles of aerotriangulation are reviewed in this is shared by three consecutive photographs (two consec-
section. The basic procedures for sequential and simul- utive stereomodels) along a strip. As the pass point is
taneous aerotriangulation adjustment methods are dis- positioned by one stereomodel, it can be used as control
cussed as background for Chapter 6. Emphasis is to orient the adjacent stereomodel. Thus, the point
placed on fully analytical aerotriangulation procedures. "passes" control along the strip. Only three-ray points
in the triple overlap area serve this pass point function.
10-20. Aerotriangulation Principles A two-ray point in the stereomodel is an intersection
point only and does not contribute to the aerotriangula-
a. Definition. Aerotriangulation is the simultaneous tion function unless it is also a ground control point. A
space resection and space intersection of image rays pass point that is shared between two adjacent strips is
recorded by an aerial mapping camera. The spatial called a tie point.
direction of each image ray is determined by projecting
the ray from the front nodal point of the camera lens 10-21. Aerotriangulation Methods
through the image on the positive photograph. Conju-
gate image rays projected from two or more overlapping Aerotriangulation methods can be characterized in
photographs intersect at the common ground points to several ways:
define the three-dimensional space coordinates of each
point. The entire assembly of image rays is fit to a. Photogrammetric projection method (analogic or
known ground control points in an adjustment process. analytical).
Thus, when the adjustment is complete, ground coordi-
nates of unknown ground points are determined by the b. Strip or block formation and adjustment method
intersection of adjusted image rays. Simultaneously (sequential or simultaneous).
with the ground point intersections, the exterior orienta-
tion of each photograph is determined by image ray c. Basic unit of adjustment (strip, stereomodel, or
resection through the camera. image rays).
10-15
EM 1110-1-1000
31 Mar 93
Modern instruments and adjustment methods dictate that propagated down the strip causing the horizontal and
an analytical adjustment of stereomodels or image rays vertical strip coordinate datums to bow and twist. Al-
be emphasized. though it is difficult to measure and completely remove
individual systematic errors in each stereomodel, the
a. Semianalytical aerotriangulation. Semianalytical total effect of all the errors acting on a strip unit is
methods use the stereomodel as the basic unit of the readily apparent after initially leveling and scaling the
aerotriangulation process. Stereomodels may be formed assembled strip unit. Discrepancies at the horizontal
by either analogical or analytical methods. After and vertical control coordinates are calculated by taking
interior and relative orientation of each stereopair of the coordinate differences between the known position
photographs, stereomodel coordinates are measured in of each control point and its position in the assembled
an arbitrary coordinate system defined in the model strip unit.
space. For aerotriangulation purposes, the spatial
coordinates of the two exposure stations and each pass (2) Strip adjustment. The sequential strip formation
point in the stereomodel are required. causes the X, Y, and Z coordinate discrepancies at the
control points to form smooth error curves. The error
(1) Sequential strip formation. A strip unit is curves can be approximated by second- or third-order
formed sequentially from the stereomodels by joining polynomial functions. Corrections to the strip pass point
them at the exposure station and pass points shared by coordinates are derived by evaluating the polynomial
the new stereomodel and the previous stereomodel functions at the strip location of each pass point. The
already in the strip coordinate system. Each stereo- corrections are applied to the original strip coordinates
model added to the strip provides the necessary points to complete the strip adjustment process.
to bring the succeeding stereomodel into the strip unit.
The sequential nature of the strip formation procedure (a) The polynomial functions model the error propa-
allows systematic errors to accumulate as the strip unit gation in a single strip unit. Block configurations may
is assembled from the individual stereomodels. Resid- be adjusted by using separate polynomials for each strip.
ual systematic errors in the interior orientation, the rela- The unique strip units in a block are joined together at
tive orientation, and the measurement system are tie points in the overlap area between strips. The
10-16
EM 1110-1-1000
31 Mar 93
polynomial functions adjusting adjacent strips are the aerial photographic images. Since the photographic
constrained to yield the same adjusted coordinates at image will have scale variation caused by the perspec-
each tie point in the block. tive view of the camera, photo tilt, and unequal flying
heights, each product must be evaluated according to
(b) The polynomial adjustment of aerotriangulation how the scale variation problem is treated. Scale varia-
is an approximate method that is less accurate than the tion may be ignored, as in an index mosaic, and an
simultaneous adjustment of stereomodels or photo- approximate map substitute is obtained; or the scale
graphs. The purpose of polynomial strip adjustment in variation can be compensated, as in an orthophoto pro-
modern aerotriangulation systems is to perform a pre- cess, and a nearly true map substitute can be obtained.
liminary adjustment of all the photogrammetric and
ground survey measurements. The sequential strip 10-23. Aerial Mosaics
formation and polynomial adjustment is an excellent
way to detect blunders in the data and provide initial Aerial mosaics are constructed from sets of individual
approximations for use in simultaneous adjustments. adjoining aerial photographs. Typically, the outer edge
of the photo coverage of each print is trimmed back to a
b. Simultaneous adjustment of stereomodels. The selected match line, and the photos assembled by care-
adjustment of independent stereomodels as individual fully matching ground detail along the match line. A
units is the logical extension of the strip unit adjustment photo index mosaic is a rough composite of a number of
discussed in the previous paragraph. Sequentially individual photographs of a flight line or set of flight
formed strip units are not used, but rather all lines overlaid one on top of the other without trimming
stereomodels in an entire strip or block are simultane- the photo prints. This composite may itself be photo-
ously transformed into the ground coordinate system. graphed onto a single piece of film. Mosaic and index
Stereomodels without ground control points are carried assemblies form a continuous representation of the ter-
into the ground coordinate system by the pass points rain covered by the photography. Photo maps are maps
and exposure stations shared with adjacent stereomodels. using a photograph (or mosaic) as the base to which
The simultaneous adjustment of stereomodels can rival limited cartographic detail such as names, route num-
the simultaneous adjustment of photographs if the indi- bers, etc., are added. Generally, a photo map is con-
vidual stereomodels are corrected for all systematic structed from an orthophoto, or from several
deformations. However, the development of analytical orthophotos mosaicked together.
stereoplotters that measure precise photocoordinates
warrants that the more rigorous fully analytical 10-24. Uses of Mosaics
aerotriangulation be required for all large-scale and
high-accuracy mapping. The increased need for presenting a pictorial view of the
earths surface has led to the aerial mosaic as a means
c. Fully analytical aerotriangulation. Fully analyti- of showing a complete view of large areas. Because a
cal aerotriangulation is a simultaneous adjustment solu- single photo is limited in area, groups of photos are
tion of collinearity equations representing all the image combined into mosaics to provide the aerial picture.
rays in a strip or block of photography. The adjustment Mosaics are of principal use for presenting synoptic
is often referred to as the "bundle" method. Although views of a relatively large area, and as indexes of indi-
the basic unit of adjustment is actually the individual vidual photographs. The features are usually labeled or
image ray, the solution determines the exterior orienta- "annotated" to facilitate the recognition of critical areas,
tion parameters of the bundle of image rays recorded by or overlays may be prepared to show planning develop-
each photograph, and the adjusted ground coordinates of ments or contemplated changes in an area. For
each ground intersection point in the ground coordinate example, a mosaic can show the location of a proposed
system. highway. Using this technique, laypersons, and particu-
larly property owners adjacent to the proposed highway,
Section V can visualize its effect on their holdings. Photo indexes
Map Substitutes Mosaics, Photo Maps, and are used as a visual record of the photographs of a
Orthophotographs project and as a reference of availability of photographs
over a particular area of terrain. Photo maps are
10-22. General
particularly useful for land use, land cover delineation,
land planning, zoning, tax maps, and preliminary engi-
Map substitute refers to those products where the com-
neering design.
piled planimetric line map is replaced or augmented by
10-17
EM 1110-1-1000
31 Mar 93
10-25. Methods of Mosaic Construction photographs. A controlled mosaic or air photo plan is
prepared using photographs that have been rectified to
There are three principal construction methods for the an equal scale, while an uncontrolled mosaic is prepared
making of mosaics: by a "best fit" match of a series of individual
photographs.
a. From paper prints. This is the simplest method.
The paper prints that are to be joined are laid one on 10-27. Orthophotographs
top of the other so that commonly imaged terrain is
matched. Then the paper prints are cut, along a feature Orthophotographs are photographs constructed from
if possible, so that the edge joins will be as inconspicu- vertical or near-vertical aerial photographs, such that the
ous as possible. The photographs are matched and effects of central perspective, relief displacement, and
pasted to a base. As necessary, the photographs are tilt are (practically) removed. The resulting
stretched to assure a good match of features across the orthophotograph is an orthographic project.
join line. The completed mosaic may be photographed
if a reduced scale mosaic, such as an index mosaic, is a. Orthophotomaps. Orthophoto maps are
desired. orthophotographs with overlaid line map data. The
common line data overlays include grids, property lines,
b. From negatives or diapositives. This method is political boundaries, geographic names, and other
called photomechanical mosaicking. The negatives to be selected cultural features.
used for the construction are overlaid and stud regis-
tered. After the join lines are selected, a mask is made b. Limitation of orthophotographs. The original
that is opaque on one side of the join line and clear on aerial negative from which the orthophotograph is made
the other. A negative of the mask is made for the other is a central projection and, as such, displays relief dis-
photo of the pair to be joined. These masks are also placement and obscuration of features. For example, a
stud registered. After all negatives and masks have building will obscure the terrain that lies behind it. This
been stud registered, a piece of unexposed film is put obscuration results in gaps of information that can be
down (in the darkroom) over the studs and the first gotten only from other sources of information, such as
photo negative and its corresponding mask are registered field survey or separate photograph. Also, optical
and an exposure made. The mask permits only that orthorectifying devices have a finite width and length
portion of the unexposed film under the negative to be slot width that can "rectify" only to an average elevation
exposed. Successive portions of the film are exposed as within that slot. Thus, small objects, such as trees,
this process is repeated for each negative and mask until buildings, etc., are not rectified at all. Normally, only a
the whole film is exposed as a single mosaic. "smoothed" approximation to the actual terrain is recti-
fied. Digital rectifiers with spot diameters as small as 5
c. From digital imagery. Digital (or digitized) to 25 micrometers can, to a much greater degree, over-
imagery can be used to create a mosaic. This method come this latter shortcoming.
has been used most commonly with small-scale satellite
photographs. The method is very computer intensive. 10-28. Uses of Orthophotographs
Either by human or automated correlation, common
points on the two images to be joined are selected. The Orthophotos and orthophoto maps are widely used for
computer can then match each pixel of the one image general planning purposes. The complete and accurate
with the corresponding pixel of the second image. A display of all features in a project area is an ideal medi-
join line is decided upon and the two images digitally um for demonstrating terrain features to laypersons.
"sheared" along the join line. The two sheared images Proposed designs of engineering projects can be super-
are then brought together into a single file having a imposed on the orthophoto map for a vivid understand-
common pixel coordinate system. ing of work to be accomplished. Many factors must be
considered before deciding on a photomap product in-
10-26. Types of Mosaics and Air Photo Plan stead of a topographical map product. The intended use
Maps of the product needs to be classified exactly, together
with the accuracy required and whether height informa-
Mosaics or air photo plans may be uncontrolled, semi- tion is required or not. For many USACE engineering,
uncontrolled, or controlled. They may be constructed operations, and maintenance activities, rectified air
from unrectified, rectified, or differentially rectified
10-18
EM 1110-1-1000
31 Mar 93
photo plan sheets may be used in lieu of fully rectified the instrument in the conventional manner for standard
orthophotographs. stereo photogrammetric mapping. The requirements for
ground control or control to be established by aerotrian-
10-29. Methods of Orthophotograph gulation are the same as for photogrammetric mapping.
Preparation The instrument provides the means of scanning the
stereomodel to effect corrections for the varying scales
Orthophotographs are prepared from pairs of overlap- caused by topographic relief. The tilt and other
ping aerial photographs using specially designed ortho- distortions are corrected in the orientation of the
plotting instruments. The photographs are oriented in stereomodel.
10-19