PaALAaUELCLD 
  
TABLE 4 
D FOR WEIC 
E 
z 
FACTORS US 
HS 
D 
SIMULTANEOUS ADJUSTMENT CF PHOTOGRAI 
11 
E. TEST RESULTS 
  
Tests using specified control arrays are labeled as follows: 
Control Point Test Case Plate Coordinate 
Array Label + Perturbations 
A 1A 
1B Random Normal Deviates, only 
(0 = 6 micrometers) 
C 1C 
A 2A Random Normal Deviates 
à 22 (0 = 6 micrometers) 
plus residual systematic 
C oC deviations. 
ve 
  
1. Results with Specified Tests 
Test results obtained by participants using sequential and simultaneous 
procedures with plate coordinates perturbed by random normal deviates only 
(Test Cases 1A, 1B, 1C) are tabulated in Tables 1A, 1B, and 1C. These tables 
display: (a) root mean square errors (RMSE)* of discrepancies in check points 
for X and Y coordinates (my, my), position (my y)» and elevation (my); (b) degree 
of equation used in block adjustment by TRA methods; (c) number of 
iterations required for convergence; and (d) number of points in the sample 
of check values. Table 1A' lists results obtained by Participant 1 with 
control array A' illustrated in Figure 4. 
Tests results with plate coordinates perturbed by random plus residual 
systematic deviations (Test Cases 2A, 2B, 2C) are listed in Tables 2A, 25 
and 2C, respectively.. Results obtained by Participant 1 for Test Case 2A 
are given in Table 24'. 
The tables cited above enable comparison of one procedure versus another 
for the same test array. To facilitate comparison of the effects of adding 
ground control and of introducing systematic perturbations in plate coordinates, 
the RMSE's for discrepancies in elevation, mz, and position, myy, in check 
points are listed for each participant for all test cases in Table 5. The 
RMSE's of discrepancies in position and elevation for expauure station co- 
ordinates are tabulated in Table 6 for all test cases. 
Additional items of interest output from direct simultaneous procedures 
are the estimated standard deviations of unit weight, mg, in plate coordinate 
residuals for the respective adjustments, listed in Table 7. Using the 
standard deviation of 6 micrometers for the applied perturbations, variance 
  
E 
* RMSE (root mean square error) - 1 where NET. calculated value - true 
value and i = 1, 2, ..., n the MEER of cech points.