On one of tre mo ste 
'ared moving b % 
TTOTS, the mode] ig loy f 
pect to the loca] Vertica] à 
ng (selecting passage Im 
nventional Methods of is 
e evident; theiy Value s 
caméra contains in Itself 
S to be made parallg ts " 
ess of the recording of 
id in the matte of ety 
on of which is accomplie 
of the latter Ones, © ani ; 
g of the nadiral point, n 
through collimations on b 
9 the Setting of the data : 
'Scopie nadiral point is ui 
r instrumental Operations y 
oh is referred to its verti] 
idging is eliminated; dy; 
in this regard it is, in fag 
rrection to the tp of em 
ertical lines passing thong 
1 
= 
+ 
i 
M 
a 
es the chaining of the phy; 
uces various manufac 
stics appropriate for pho 
Vy, to their rigorous dem 
broader liberties with rui 
with types of gyms 
lon of the indications of 
1ass and the relatively lv 
the results of the laboratory 
cpected from the complex di 
9jeyond the laboratory tes, 
to give any concrete real: 
nditions for the purpose if 
the “European Organiza 
experiment with the Ni 
f linked aerial photogr 
titude with the OMINs: 
ments with these expo 
ributed the material to 
e of concordance with who 
od. V or the Nistri Pit 
or rather the nadiral pit 
instruments this opertit 
ary rotation of the camer 
5 
. axes of the two gyroscopes, equipped with erectors, operate under extremely 
ENS dits the average difference of 670 recordings given by the two gyros- 
ugh fie e two hours flying time was + 5’. Other experiments of the OMI, limited 
ae ie and not comparable to those in the OEEPE project, have given an 
however in attern of extraordinarily good results for short stretches of the photographic 
itemating m less good results; it is too difficult, with what is yet a too small number 
dai pal to separate the various causes of such a performance. 
en eius determined, nonetheless, that beyond the need to create gyrocopes parti- 
jy suited for this work, as has already been said, it is necessary that all provisions 
ary designed to limit to the maximum the spreading of the accelerations of the 
le ma t of the aircraft to the recording system, or to the take camera, that is, as is 
ET installation of the cameras stabilized by gyroscopic systems, already 
ig with elsewhere, in aircraft equipped with an automatic pilot and radio- 
m level flight controls, would assure ideal working conditions for the Nistri gyroscopic 
T. point recording system. 
While the absolute precision of the method cannot yet be expressed in figures, one 
a instead already state precisely the nature of the error which effects the gyroscopic 
uli] point recorded by the Nistri method; it is completely accidental, if considered, 
sit must be, discontinuously, photograph for photograph. We must in fact exclude a 
ine phase between the release intervals of the aerial takes and the deviations, mostly 
sal, on the order of 2, that the gyroscopes present in static conditions in their housing; 
ih regard then to the accelerations to which the gyroscopes are subject during the aerial 
apsures, one can deduce that only an eventual curvature in the line of flight, as obtained 
iylinking the exposure points with a continuous line, can introduce errors of a systematic 
mein the indications of the gyroscopes, extended to the whole chain; but, such curvature 
king the very slightest, because of the characteristics of a good picture-taking flight, this 
copes in 
if tests aval 
wording of the nadir position. 
Having considered this, let us deduce how to act when, at the end of the simple operation 
if chaining originally described, the finished model presents y-parallaxes superior to a 
uitain tolerance, as fixed in accordance with the type of plotter used, or it presents obvious 
mors of orientation. In such a case the gyroscopic nadiral points can be used only for the 
moses of absolute orientation, but perfecting the model along with it by clearing the 
Haallaxes with the conventional methods of relative orientation. The operation is the 
filoving: the new photograph oriented, having recourse to its gyroscopic nadiral point, 
i11", k" and b," the distribution and the entity of the y-parallaxes in the model must be 
wsdered; one will thus deduce the variations of tip and tilt necessary to perfect the 
itive orientation. Such variations will be applied to the new photograph only in the 
went that they are large and that the preceding model had been found free from residual 
allaxes; otherwise the said variations can be applied, half to the precediny photograph 
aibalf to the new one. By this we will have accepted an absolute orientation of the model 
#given by the average of the gyroscopic indications of the photographs of the pair con- 
ered, thus always keeping in mind the accidental character of the error in such in- 
ations, Obviously, every time you have to resort to such an operation. altimetric 
"smtinuities will be noticed in the Strip, to be corrected by the act of compensation, while 
t Nec: discontinuities will remain in every case an order of size that can be 
erlooked, 
ai from the ever larger approximations of which the gyroscopes will be capable, 
MAN Im it is Interesting to consider the possibility described of using the gyroscopic 
B NUN vm bridging, also for the sole purpose of absolute orientation. For every 
UM an eon deviate from the true one, but it will always exist in small quantities, 
i. y s ributed around the vertical locus, with the effect of limiting the spreading 
OS normally having a systematic character. 
  
  
aie is not sufficient to diminish the accidental nature of the error in the gyroscopic 
  
MEME It afm ee re PR 
  
  
  
  
  
  
  
  
es