Full text: Reports and invited papers (Part 3)

The collinearity equations (or its distance equivalent) be used spatial networks 
for points addressable at the same time j 
A simple and strict space intersection of corresponding points is therefore only possible if all 
orientation elements of two overlapping fliehts are known as functions of time for the two corres- 
ponding image points. Tor this reason the 6 orientation parameters x_ y. Lo zw, $ ,K 
should be determined by navigational and inertial devices fo ach imag ‘light with an accuracy 
J 
corresponding to image resolution. 
The differential formulae (1) and (5) show the approximate effects of the orientation parameters 
on terrain displacements. It is of special interes that dx , 1K anc ad only have an influen 
- o a 
on dx. , dy, 5 dz, as well &s do only influence dy. (for radar images d does not have 
Yo 3 1 
+ 
o2 
furthermore it is to be noted tha 
t d$ and dx 'e comietely correlated 
i 0 
dynamic image acquisition systems the 
velocity v' C varies greatly with the groundspeed of the platform 
V and the chosen image scale 
c 
y! = 
Z 
This means that small scale images will show high 
images will have lower frequency oxillations for the same flight path and 
variations. 
If stereocoverage by two overlapping scans is not practicable (or by complications 
tution tchnology does not become desirable) then restitution only becomes possible if the 
imaged object constitutes & known surface. Such a surface may be of & simple nature (plane, 
sphere, ellipsoid) or it may be described by & complicated relationship in form of some type 
of digital terrain model. Each image element (pixel) must be associated with a corresponding 
elevation 2; j then the solution for Xi and Y; becomes possible by the ground - coordinate 
form of the collinearity equations derived from: 
B43: 2123 
2213 2223 
2315 2323 
  
  
The problem then only consists in expressing the surface or DTM information as functions 
image coordinates 
= f (xi yi) =f (xi 01) 
For active (radar) images an analogous approach is possible: 
For overlapping images a bisection of the two spheres défined by the slant ranges of the corres- 
ponding images as radii may be made. The resulting circle may be bisected with the antenna plane 
or come to yield one out of two point solutions separable by the condition (2, - 2.) 0 
Clerici recently formulated and tested this problem in yet published form. If Zi can likewise 
be expressed by (74) for & D.T.M. then the circle defined by t radius representing yi located 
in the plane (as expressed by (2a) or on its equivalent may be intersected with the line z. 
given by the D.T.M. for y1 . This will determine X. and y. . 
L À 
 
	        
Waiting...

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.