×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
Proceedings, XXth congress

International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B-YF. Istanbul 2004

Figure 3. The panchromatic orthoimage using control
points measured with GPS.
4. ACCURACY ESTIMATION OF THE TWO
ORTHOIMAGES
4.1. Required accuracy from the Hellenic Cadastre Project
As it is known the technical specifications of the Hellenic
Cadastre fixes bearable limits for horizontal and vertical
accuracy rise up to two (Technical specifications of Hellenic
Cadastre Project, 1999) (Arvanitis A., 2000).
In order to achieve accuracy control the technical specifications
demand the coordinates’ comparison of certain characteristic
points in the orthoimage with the coordinates that were received
with any method (digitalization, GPS).
The map accuracy is expressed by variance o^, or the Root
Mean Square error of the points' coordinates, that was
calculated (Patias P., 2001):
e From the n points' horizontal coordinates

5 ] 2 ~ 5 ~ ”
e e o tns
2n t7 (1)
1 $ > / >
2 Ó C
Gr = & = x, ) T (v; "xdi ¥
2n ^7
e From the n points heights
Î n 2
2 Ó
f —- = 2
o ver = S. (z i z i ) (2)
n ja
Where
X ; y ? s : = the values of control points’
coordinates in the orthoimage
X. Vi ZZ, Me corresponding topographic control
1
values.

O0 2 : A :
Ox; ; Óy ; * the coordinates" residuals
N= the number of points for the accuracy control
2 2
Q noc — the variances' estimation for the horizontal
> ver
and vertical control
Before the value o becomes acceptable, the values dxi, dyi are
checked for probably gross errors applying data snooping
control:
ÓX;, y; « 1.96 Ohor (3)
02; $2.16 0. (4)
The percentage of the rejected values resultant from the
application of this control should not exceed the 5% of total
values. After the rejection of the off limits deviations the root
mean square error of the remaining points is calculated and the
global test takes place. The value o is checked with the help of
the reference variance. The prices o,, which come in this case
from the technical specifications of the Hellenic Cadastre (oy,
Kat 607), depending on the scale, are given for the coordinates x,
yz.
e For horizontal accuracy:
Gun 0 9264 22 096,0 (5)
x
e For vertical accuracy:
Lnd
Over £0980, — 0.96, 1 (6)
Accordingly to the statistics if the inequalities are valid, the
estimation of accuracy that was calculated is oxy-2m and
o07-2m by tacking into consideration the technical specifications
of the Hellenic Cadastre about o, as well.
4.2. Orthoimage's accuracy estimation that was produced
with the method of digitization
Nine points were used for the accuracy control, which were
distributed in the total of region. The points' coordinates were
received from ortoimage and they were compared with the
coordinates that were received by the digitization of surveying
diagram in scale 1:5000. After calculations it was calculated
that 6,,,=1,245m and 6,.,=0.412m for the nine points.
The data snooping control for gross errors in the nine control
points accomplished successfully. There was no necessity to
reject any point because of the type:
0X;,0y;<1.960},,,=2.44m (7)
As well as the type:
02;<2.160,,=0.89m (8)
Continuously a total accuracy control was applied, where the
following type should be verified:
Gu 50.92 exy cx 0.9 oy (9)
As well as the type
0.20.9980, 2 090, (10)