Image displacement due to tilt
Considering point, a in Figure 1, whose distance from the isocentre on the tilted
photograph is ia, and on the un-tilted photograph is ia1, then the displacement ∆t due to
tilt is given by
From consideration of the geometry of Figure 1 it can be shown that
Figure 1 Un-tilted photograph
where y is the distance from the isocentre measured along the principal line in the
direction of downward slope.
For any point off the principal line, the situation is as shown in Figure 2.
Page 1 of 6
Notes by Sichangi Arthur
Figure 2 Tilt and image displacement
so, the general equation becomes:
The equation shows that displacement is proportional to the distance from the isocentre
squared, and will therefore be greatest at the edges of the photograph. It shows also
that increasing the focal length of the camera will help to reduce the displacement. It
can be shown that image displacement due to tilt must be radial from the isocentre.
In addition to the complications already outlined, further displacements may result due
to variation in flying height, refraction of the rays of light (particularly near the body of
the aircraft), camera and photographic errors, etc. It can now be clearly seen that a
photograph is not a plan, except where the axis of the photograph is truly vertical and
the ground is flat and level.
Page 2 of 6
Notes by Sichangi Arthur
Combined effect of tilt and relief
As already shown, displacement due to tilt and relief are not radial from any one point
on the photograph.
Figure 3 shows ab as the top and bottom of a tall structure. The height displacement ab is
radial from the plumb point v, whilst the tilt displacements aa1 and bb1 are radial from the
isocentre. Note the reverse direction of displacement on the upper side of the photograph.
Figure 3 Effect of tilt and relief
The treatment of such effects is to:
i. Eliminate tilt displacement by a mathematical or optical rectification of the
photograph, i.e. reduce the tilted photograph to its horizontal equivalent;
ii. Consider height displacement on the rectified photograph – for instance, after
rectification the equivalent position of the plumb point v is at v1 (see also Figure 1)
from which the height displacement a1b1 is radial.
To find the x and y tilts of a photograph
Knowing the focal length of the camera and the coordinates of the plumb point, the x and
y tilts are given by
Page 3 of 6
Notes by Sichangi Arthur
where θ is the tilt of the photograph and θy and θx are the component parts of the tilt in
the y and x directions respectively. Similarly yv and xv are the component parts of the
vector pv in the y and x directions. Therefore, all lines parallel to the principal line have
the same maximum tilt and all lines at 90◦ to the principal line (i.e. plate parallels) are
horizontal.
Ground coordinates from a tilted photograph of flat terrain
In Figure 4 consider point a whose photo coordinates xa and ya are measured about the
fiducial axes.
Figure 4 Tilted photograph of flat terrain
The flying height is H, the focal length is f, the tilt is θ and the swing is S. It is first
necessary to obtain the coordinates of a relative to the principal line axes, with the
isocentre as origin (tilt displacement radial from isocentre), i.e. x’a, y’a. Figure 4(b)
illustrates the rotational effects where the angle between the respective axes is α = 180◦ +
S. The amount of translation necessary is pi = f tan(½θ). Assuming a parallel through p,
an examination of Figure 4(a) shows that the angle between the x-axes is a. Then
from Figure 4(b):
Page 4 of 6
Notes by Sichangi Arthur
To obtain the general form for these equations, substitute (180°+S) for α and add the
translation amount pi = f tan(½θ) to obtain the new origin at i, then
where x and y are the photo coordinates measured about the fiducial axes.
The equation, for scale on a tilted photograph ( ), can now be applied to
the reduced coordinates to give the ground coordinates X and Y as follows:
Ground coordinates from a tilted photograph of rugged terrain
As the effect of ground relief is radial from the plumb point, then the rotation and
translation is this time relative to v, where pv = f tan θ, then:
Thus, from Figure 5, the new coordinates of a are:
Page 5 of 6
Notes by Sichangi Arthur
Figure 5 Tilted photograph of rugged terrain
Mathematical rectification for each point is now carried out by considering horizontal
planes passing through the plate parallels of the points in question. For example, in the
case of point a, the horizontal plane through it is qra, and the rectified coordinates are
therefore
The new focal length appropriate to the plane of rectification is now
We now have an un-tilted photograph of ground point A with a new focal length f1. It
now only remains to multiply the photo coordinates by their appropriate scale as giving
Page 6 of 6
Notes by Sichangi Arthur