International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B2. Istanbul 2004
laser point
h=
height
w = width
„nn
a
| = length
Figure 2. Soil volume determination for construction works in
the floodplains.
precision (for | m digging height) for areas of about minimal 50
ha in case of no vegetation. This applies for point densities of
15 and 30 pt/m°. The disturbing effect of (low) vegetation on
laser measurements still is difficult to quantify but is estimated
to be 2-3 cm (systematic error). For areas with vegetation, the
FLI-MAP system is satisfactory for digging depths from
minimal 2 m. Laser data acquired from an airplane with a point
density of about 1 point per 16 m? can be used for digging
depths from minimal 2 m in case of no vegetation and minimal
3 m in case of vegetation. Concerning the costs, helicopter
borne laser altimetry is about 30% cheaper than terrestrial
measurements for small areas (50 ha) and about 50% for larger
areas (200 ha). This applies for an extreme high point density of
30 pt/m^. Another advantage of laser altimetry is the possibility
to survey large areas in short time.
4. GROYNES
Along the river banks of the large Dutch rivers we find
numerous groynes (see fig. 3). These are structures of sand and
stone extending from the river bank into the water, built
transverse to the flow direction. Their main purpose is flow
acceleration to minimize sediment deposits in the main channel
(fairway). This way the required depth of the riverbed for
shipping is maintained. Monitoring of groyne shapes is
important because different kinds of damage may occur which
affect the performance of the groynes, for example collisions
with ships and deformations by the current.
MS II 5
Figure 3. Groynes
The present way of monitoring the above-water parts of
groynes is with terrestrial measurements and visual inspections.
The measurements consist of levelling profiles (one profile
across the ridge of the groyne and two profiles perpendicular to
the main axis) and slope measurements with a 3 m rod. In
addition, visual inspections are performed for damage control
and acquisition of the number and approximate heights of trees
and bushes growing on groynes. In fact, groynes must be free of
(high) vegetation. These activities are time-consuming and
therefore expensive. Thus they are carried out not more than
once per three years. The below-water parts are measured with a
multi-beam echo-sounder system.
Concerning costs and acquisition time laser altimetry seems to
be a good alternative for the terrestrial groyne measurements.
The main question, however, was: what is the precision of the
laser altimetry data and could groyne deformations be measured
with laser altimetry? Therefore, a test was performed with the
FLI-MAP system. This is a helicopter borne laser scanner
which acquires laser altimetry data with extreme high point
densities: 15-20 points per nr. [n our test the point density was
even higher: 30-40 points per m?. This could be achieved by
flying all strips two times. In addition, the FLI-MAP system
acquired video data (pixel size 30 x 30 cm?) and digital
photographs (pixel size 6 x 6 cm”) during the flight. Figure 4
gives an impression of the dense laser data and to what exter
morphologic details can be recognized. Apparently, there ar
very few laser reflections from the water surface (black pixels).
L1
Figure 4. Hillshade of laser data grid of a typical groyne with
Lits . : ^ 2
some bushes on it (grid cell size: 20 x 20 em”).
The laser data were compared with dense GPS measurements
on four groynes (about 200-400 measurements per groyne) with
varying grid cell size and point density. The main conclusions
of this comparison were:
e The height differences showed standard deviations of
about 7 cm. This contains not only the laser scanner
point noise (in this case about 4.5 cm), but also the
error of the GPS measurements and the roughness of
the groyne surface.
e'The systematic error lay between -0.1 cm and -7 cm.
Several centimetres height shift of laser strips can be
caused by positioning errors of the helicopter.
e The extreme high point density of 30-40 points per m' is
not necessary: 15-20 points per m^ would be sufficient
and does not affect the achieved accuracy.
e A grid cell size of 50 cm is recommended because of the
roughness of the groyne surface and the averaging of
Do
laser point noise (several points per grid cell).
e The laser data is not suitable for monitoring deformations
per single grid cell. The deformations must be
considered for at least some neighbouring grid cells or
the whole groyne surface.
e Absolute deformations can be detected from about 10 cm,
relative deformations (comparing one laser data set
versus another) even from about 5 cm.
Considering the costs of terrestrial measurements even about
40% of the total costs can be saved using laser altimetry. This
applies for a point density of 15-20 points per m.
236
Interna
In addi
groynes
of flow
rough |
informa
because
that the
heights
(15-20
acquire
probabl
tests on
Wherea
can be
below-\
sounder
monito!
water I
whole |
floodpl:
datasets
measur
laser da
gaps w
account
models.
y [m]
42500
42495
42490
42485
42480
42475
42470.
Figure :
multi-be
Laser a
and ech
Structur
an anim
this ani
innovat
visualiz
the. futu
technic:
Usually
beam e
floodpl:
level. T
zone, bi
the gro
measurc
