lume XXXIX-B3, 2012
nd. Thus, this mentioned re-
sible. Bundler automatically
iation based on a small num-
mage size). It detects natural
ow one pixel in image space
est applications as well. The
id demonstrate that the post
le. PMVS provides a patch-
ts from Bundler to create a
Based on a extended multi-
ases the number of points to
ith given image sequence of
] orientation and computed 3D
ise point cloud
ller and PMVS2
ccuracy is comparable with
.1). However, illustrated re-
one image pair. Because of
/ of all following images in-
| values (w,®,K) are worst in
cause for this effect is a clus-
ages, homologous points are
in only one region of the im-
rs can not be computed with
ppagation, successive image
Detected and stored feature
ive orientation between two
ling images could be linked
se estimation. One possible
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
| parameter I mean standard deviation for 36 frames ]
0.04mm
0.13mm
0.04°
0.19°
0.06°
Table 4: Mean accuracy of relative orientation between one im-
age pair
x [SE [N|=
implementation is discussed in (Davison, 2003) and (Davison et
al., 2007). It is an advanced model of VSLAM with an Extended
Kalman Filter (EKF). At the moment, our work does not support
a prediction and SLAM algorithm. Nevertheless first results and
existing works demonstrate the potential of this method, which
we strive to develop further.
6 CONCLUSIONS AND FUTURE WORK
First results of this pilot study show the feasibility of theoretical
background and it is worthwhile to pursue this approach further.
A vision based method can estimates a position of a forest har-
vester with an adequate accuracy. Especially the post-processing
approach delivers stable and precise results. The Pose estima-
tion results below one cm in experiments using post and real time
processing exceed the expectations and form the basis for fur-
ther research. Furthermore, the achieved results meet accuracy
requirements of about 15cm, i.e. half of a tire width, for for-
est navigation to avoid soil compression. Influences on accuracy,
such as strong wind and fast movements, have not been analyzed
yet. The calculated point cloud of Bundler and PMVS2 can be
used to create a virtual forest. Current approaches for automatic
tree detection ((Heurich and Weinacker, 2004),(Schilling et al.,
2011)) for laser scanner data can be applied on this data. In this
context, another possible navigational aid could be a virtual for-
est model in which cylinders have been fitted to appropriate point
cloud subregions. Assuming all trees of this virtual forest have at
least 2D coordinates, a camera on a harvester and a corresponding
software have to solve the assignment of trees in object and image
space and could compute their position by spatial resection.
REFERENCES
Bay, H., Tuytelaars, T. and van Gool, L., 2006. Surf: Speeded up
robust features. Technical report, ETH Zurich.
Davison, A., 2003. Real-time simultaneous localisation and map-
ping with a single camera. In: Proc. International Conference on
Computer Vision, Nice.
Davison, A. J., Reid, I. D., Molton, N. D. and Stasse, O., 2007.
Monoslam: Real-time single camera slam. IEEE Transactions on
Pattern Analysis and Machine Intelligence 29, pp. 1052-1067.
Fischler, M. A. and Bolles, R. C., 1981. Random sample con-
sensus: A paradigm for model fitting with applications to im-
age analysis and automated cartography. Communications of the
ACM 24(6), pp. 381—395.
Furukawa, Y. 2010. Patch-based multi-
view stereo software (pmvs - version 2).
http://grail.cs.washington.edu/software/pmvs/.
Furukawa, Y. and Ponce, J., 2010. Accurate, dense, and robust
multi-view stereopsis. IEEE Trans. on Pattern Analysis and Ma-
chine Intelligence 32(8), pp. 1362-1376.
61
Hamberger, J., 2001. GPS als Mittel zum umweltschonenden
Maschineneinsatz: Navigation von Forstmaschinen und Doku-
mentation ihrer Fahrbewegung. PhD thesis, Technische Univer-
sitt Mnchen.
Heurich, M. and Weinacker, H., 2004. Automated tree detec-
tion and measurement in temperate forests of central europe us-
ing laserscanning data. Int. Arch. Photogramm. Remote Sens.
XXXVI, pp. 198-203.
Lourakis, M. A. and Argyros, A., 2009. Sba: A software package
for generic sparse bundle adjustment. ACM Trans. Math. Soft-
ware 36(1), pp. 1-30.
Lowe, D. G., 2004. Distinctive image features from scale-
invariant keypoints. International Journal of Computer Vision 60,
pp. 91-110.
Schilling, A., Schmidt, A. and Maas, H.-G., 2011. Automatic
tree detection and diameter estimation in terrestrial laser scanner
point clouds. In: 16th Computer Vision Winter Workshop.
Snavely, N., Seitz, S. M. and Szeliski, R., 2007. Modeling the
world from internet photo collections. International Journal of
Computer Vision 80, pp. 189-210.
