2.6 Point Cloud Registration
The spatial relation between two point clouds with n 3D/3D cor-
respondences X; €» X; with X;, X; € mR? can formally be
described as
X;=RX;+t (6)
where R. € R?*? represents a rotation matrix and t € IR? rep-
resents a translation vector. A fully automatic estimation of the
transformation parameters can be derived from minimizing the
error between the point clouds. Including a weighting w; € R for
each 3D/3D correspondence X; € X; yields an energy function
E with
E — ) w|X; - (RX; t| (7)
i=1
for the registration process. For minimizing this energy function
E, the registration is carried out by estimating the rigid trans-
formation from all 3D/3D correspondences and the weigths are
derived from a histogram-based approach. This approach is ini-
tialized by dividing the interval [0m, 1m] into n, — 100 bins of
equal size. For all detected correspondences, the calculated qual-
ity measures c; and oc; assigned to the 3D points X; and X are
mapped to the respective bins b; and b;. Points with standard
deviations greater than 1 m are mapped to the last bin. The oc-
currence of mappings to the different bins is stored in histograms
h = [h;], 100 9nd h' — [e| a "da Loo Subsequently, cu-
mulative histograms
h.= Ec " and hi = ba
j=1 j=1
are derived. The entries of the cumulative histograms reach from
0 to the number n of detected correspondences. As points with
a low standard deviation are more reliable, they should be as-
signed a higher weight. For this reason, the inverse cumulative
histograms
izl,.., 100 i=1,..., 100
Beine = ^ TE S " (8)
juil
i=l,..., 100
and
ÿ=l i=i. 100
are calculated. Finally, the weight w; of a 3D/3D correspondence
X; + X; is set to
w; = minfh. n0(0:), hEinu( 71)! (10)
where c; and c; are considered as quality measures for the re-
spective 3D points X; and X;. Estimating the transformation
parameters can thus be carried out for both range imaging de-
vices separately. However, as the relative orientation between the
devices is already known from a priori measurements and both
devices are running synchronized, the rigid transformation can be
estimated from the respective correspondences detected by both
devices between successive frames. Combining information from
both devices corresponds to extending the field of view and this
yields more reliable results for the registration process. The ex-
tension can be expressed by transforming the projected 3D points
X, which are related to the respective camera coordinate frame
(superscript c) into the body frame (superscript b) of the sensor
platform according to
XR XI +t (11)
422
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B3, 2012
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia
where R? describes the rotation and t^ denotes the translation
between the respective coordinate frames. For this, it is assumed
that the origin of the body frame is in the center between both
range imaging devices.
2.7 Object Detection and Segmentation
As 2D SIFT features have already been calculated for the reg-
istration process, they can also be utilized for detecting special
objects in the scene. This allows for calculating the coarse area
of an object and for automatically selecting features which should
not be included in the registration process as they arise from ob-
jects which are likely to be dynamic. These features have to be
treated in a different way as the static background being relevant
for registration. For this purpose, image representations of sev-
eral objects have to be stored in a database before starting the
surveillance application. One of these images contains a tem-
plate for the object present in the scene, but from a different
measurement campaign at a different place and at a different sea-
son. Due to a similar altitude, the active intensity images show a
very similar appearance. Comparing the detected SIFT features
of the normalized active intensity image to the object templates
in the database during the flight yields a maximum similarity to
the correct template. Defining a spatial transformation based on
the SIFT locations as control points, the template is transformed.
The respective area of the transformed template is then assumed
to cover the detected object. This procedure allows for detecting
both static and moving objects in the scene as well as for decou-
pling sensor and object motion. Hence, the presented approach
for registration also remains reliable in case of dynamic environ-
ments if representative objects are already known.
3 ACTIVE MULTI-VIEW RANGE IMAGING SYSTEMS
The proposed concept focuses on airborne scene monitoring with
range imaging devices. For simulating a future operational sys-
tem involving such range imaging devices fairly realistically, a
scaled test scenario has been set up. However, due to the large
payload of several kilograms for the whole system, mounting the
required components for data acquisition and data storage on an
unmanned aerial vehicle (UAV) still is impracticable. Hence,
in order to investigate the potentials of active multi-view range
imaging systems, a cable car moving along a rope is used as
sensor platform which is shown in Figure 3. The components
mounted on this platform consist of
e two range imaging devices (PMD[vision] CamCube 2.0) for
recording the data,
e anotebook with a solid state hard disk for efficiently storing
the recorded data and
e a 12 V battery with 6.5 Ah for independent power supply.
As the relative orientation of the two range imaging devices can
easily be changed, the system allows for variable multi-view op-
tions with respect to parallel, convergent or divergent data acqui-
sition geometries.
However, due to the relatively large influence of noise effects aris-
ing from the large amount of ambient radiation in comparison to
the emitted radiation as well as from multipath scattering, the uti-
lized devices only have a limited absolute range accuracy of a few
centimeters and noisy point clouds can be expected. Furthermore,
due to the measurement principle of such time-of-flight cameras,
the non-ambiguous range R, with
Hs (12)
s.
Im
N| =
Figure ?
equippe
depends
speed o
sponds t
this rang
wrappin
2012). \
has to b
and that
by choo
The est
quires t
world c
frame o
dinate f
platforn
tangent
directio
à succe:
the mat:
ing 2D/
yields 3
mation
finally,
view ar
describ
devices
Y [m]
Figure
cloud d
A limit
no refe
the pos
the rel:
project
proxim
trajectc
aerial 1
betwee
has bei
maxim
A row
