A
the distance
cted data and
data (m)
lologies to be
| municipality
udies of the
| activities is
and in many
he design of
tography and
usted as the
iunicipalities,
done in a
I on the field
e consuming.
ones used at
e toa 2.5m
2P should be
er should be
, Korgaja et
phy and for
challenging
fluence the
ted accuracy
itations, the
ility and the
lex. iis
undaries are
epending on
was chosen.
was not the
'sented only
suggest an
tral plans, as
d size of the
», peri-urban
effects and
ained more
5 000) than
resampling
al resolution
between the
the image
'aphy of the
ery low for
jn and for
s in rugged
properties
Istanbul 2004
re
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part Bl. Istanbul 2004
presented a considerable variation of the surface. Important
variation of the surface can have significant consequences in
land development planning and legal aspects.
For large extensions in flat terrains and peri-urban parcels, the
variation of the distance between the surveyed border and the
extracted data was less than the pixel resolution — 5m — of the
images from which the orthoimage was deduced. Although for
rugged terrains, the difference was less than 10m. This
variation depended on the nature of the fence, vertices under
live linear enclosure were easily identified although the ones
located under dense vegetation and large canopy were identified
with less precision. It was also the case when non-physical
boundaries existed — only agreements between neighbours or
legal invisible boundaries — as presented mainly in peri-urban
areas and parcel presenting survival cultures (corn or bean
plantations).
The variation of surface and distance also depended on the
number of vertices; in this study most of the parcels surveyed
presented squared shapes, reducing the number of vertices. The
identification on the orthoimage of parcels presenting curved
boundaries, as it is the case when the border is a river, will be
less accurate. If this condition is combined with topographical
accidents, parcels of reduced size as well as a dense vegetation
cover, the use of SPOT 5 orthoimage for cadastral purpose will
be limited.
The use of SPOT 5 orthoimages as a tool for the indirect
method for cadastral projects seamed interesting for latifundium
properties presenting low variation of topography and for
medium extensions. Nevertheless, small rural properties, peri-
urban and urban estates were not well identified presenting, an
important variation of their surface as well as a deformation of
their shape as it was the case for parcels smaller than 500nY and
urban buildings. In the case of urban centres, steep gradient
areas, regions’ presenting an elevated density of parcel and
dense vegetation fences, the use of SPOT 5 orthoimage is
limited. Even if it is not feasible for collecting regional data or
data requiring continual validation because of the cost and
logistical difficulties in tropical areas, Harvey and Hill (2000)
estimated that aerial photographies make a very useful data
source due to their textural feature and superior spatial
resolution. Still, because of the cost of acquiring data and the
time involved in analysing large data sets in aerial
photographies, where the required information can be extracted
from high resolution satellite imagery — latifundium estates for
example or extended non agricultural areas — it appears to be the
most feasibly technology to adopt in land surveying projects.
And where the use of the image presents its limitations, a
combination of direct surveying methodologies and orthoimage
identification can be made.
As noted previously, the average difference between the
identified border and the reference data varied between 3.7 and
8.5m. The tolerance varies depending on the type of cadastre,
different types exist depending on the information provided and
the use of it. For example physical cadastre considers only the
measurement of the parcels. À legal cadastre consists of te
measurement of the parcel and the legal rights of this one.
Fiscal cadastre defines the parcel as the base for land taxes.
Polyvalent cadastre or multi-purpose ones is a combination of
the distinct types of cadastre. In Guatemala where a cadastre
for legal purpose is considered, the tolerance for urban parcels
is 0.3m and for rural properties is Im. In Colombia, the
tolerance varies on the land tenure; in the Netherlands the
difference of 0.28m for urban parcel is accepted and of 0.57m
for rural properties. Wilcox (1984) noted that the rural
tolerance in the US varied between 0.1m and 0.6m, and 0.3m
for urban and suburban areas. This study demonstrated that the
difference of distance surpass the established national tolerance
205
making inadequate the we of the orthoimage for delimitation
purpose in Guatemala. Nevertheless, the use of SPOT 5
orthoimage as an input for indirect methods for land surveying
can be considered in countries having fiscal and physical
cadastre, where the accuracy of identification is not as strict as
Guatemalan cadastral norms.
6. CONCLUSION
High-resolution satellite imagery with its utility for surveying
large areas in a time and cost effective manner, was considered
as an interesting input for indirect land surveying methodology.
The principal objective of this paper was to investigate the
potential use of SPOT 5 orthoimage for cadastral surveying.
Maps (1/50 000, 1/25 000 and 1/10 000) constitute useful tools
as they bring important technical elements assisting the
established cadastral activities. For countries like Guatemala,
where planning cadastral activities are prepared with outdated
topographical maps, SPOT 5 imagery brings a significant
support to cartography.
Comparisons between data from total stations — GPS
measurements and orthoimage identification for different types
of parcels were evaluated. Good results were obtained for large
and medium extensions; although it presented its limits for
identify accurately small parcels, peri-urban and urban estates.
The accuracy of identification depends directly on the size and
shape of the property, the topography of the area, the type of
fences and vegetation coverage present on the study area as well
as the scale of the orthoimage used in the identification process.
SPOT 5 orthoimage is considered a interesting alternative
input for indirect methodology for latifundium and medium
estate with low topographical variation. However for snall
properties, urban cadastre and building identification it is
recommended to use traditional surveying methods. Even if in
this study, the difference between identified vertices and the
ones located with traditional methods surpassed the Guatemalan
cadastral tolerance, the use of SPOT 5 orthoimage as an input
can be consider in countries regarding cadastre with less strict
precision.
REFERENCES
Axes, F., 2004. Spot 5 imagery as a tool for land management
and agriculture monitoring, towards the accession to the
European Union. SPOT Image.
Bernard, M., and Munier, P., 2003. FIG Working Week
“Evolution in the use of space imagery, trends and challenges".
Paris, France.
http://www ddl.org/figtree/pub/fig 2003/PS 3/PS3 1 Bernard
Munier.pdf (accessed 20 Jan. 2004)
Campell, B., 1996. Introduction to Remote Sensing. 2" edition,
Taylor and Francis Ltd, London.
Elayachi, M., Salimi, H., Souhaili, S., 2001. FIG International
Conference New technology for a new century "GPS DAO —
application to use GPS for topographic and cadastral plans",
Seoul, Korea.
http://www ddl.org/figtree/pub/proceedings/korca/full-
papers/pdf/sessionI4/elayachi-salimi-souhaili.pdf (accessed 20
Jan. 2004)
Garcia, M., 2001. Feasibility evaluation of the SPOT satellite
for parcel boundaries identification. Thesis presented to obtain
