Natalia Moskal 
  
F 
Here Q = 3Y' E EF (12) 
) E 7 covariant matrix of measured values 
1.3 Bearing data include the following data classes: elements of projection; elements of geodetic orientation of 
photography (linear and angular); coordinates of the control points; geodetic survey ( angles, directions, length of lines, 
exceeding and so on); size and forms of objects and some others. 
A separate equation is set for each type of data. This equation later is introduced into the general system of equations 
in the process of adjustment. 
4.1-4.2. The notion of the function of losses and notion of minimization of regressive remains is widely used in the 
g y 
regressive analysis [1]. The function of losses for the linear model of regressive function is as follows: 
p (e)=F 
here € — - regressive remains, d - parameter of unsquareness (—1« d € 0). 
24d 
; (13) 
  
If d = -1, the so-called robust-method of evaluation is applied. If d = 0, we got a square function of losses or in other 
words classical method of least squares. 
The use of the notion of the function of losses provides with the chance to process the series of surveys with errors 
division different from Gaussian. 
5.1-5.2. To use of the hypothesis about dependence (independence) of measuring equitation it is necessary to know (or 
neglect) the covariant matrix of errors of measuring. On the other hand, introducing covariant matrix 3. into 
mathematical model provides for logic and correct solution. This condition is used in the present research. 
2. GENERALIZED MATHEMATIC MODEL OF PHOTOGRAMMETRY SURVEY AND ITS THEORETICAL 
SOLUTION 
In this article we are discussing the general problem of joint adjustment of the measured quantity functions, control 
data and direct measuring in the following order: 
Let us admit that there are set the following data: n-dimensional vector Y of measured quantities, which are free fron 
systematical errors; its covariant matrix X, : r - values of functions are calculated. 
1 
T =F(f,) Ql 
It is necessary to make the adjustment with respect to r-conditions. 
O=®T ¥U’)=0 (22 
here T ^,Y ',U' — correspondingly means adjusted values of functions 
T'zT + AT, Q3 
control data Y' = Y ty (24) 
  
626 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000. 
Besides 
with the 
The cov 
We con: 
Gauss. 
matrix | 
another. 
When c 
Substitu 
wi 
Let us m: 
where d 
Let us co 
symbols