number dealt specifically with WG themes of integration of
imaging technology and sensor calibration.
The second main activity of 1999 was the annual CMSC
Conference, held in Seattle from 26th-30th July. As a result
of this meeting, ISPRS commenced a process by which a
formal relationship with the CMSC could be established, in
recognition of the prominent role played by the technology
of close-range digital photogrammetry in the field of indus-
trial measurement. CMSC’99 incorporated two technical
sessions specifically for presentations related to ISPRS
WG V/1 activities.
Activities in 2000 concentrated upon supporting Commis-
sion V in its organisation of the scientific programme for
the Amsterdam ISPRS Congress, where WG V1 has two
main technical sessions.
WG V/2 "Integration of Photogrammetric Systems with
CAD/CAM"
by Chairperson: Jürgen Peipe, Bundeswehr
University Munich, Germany
Stuart Robson, City University,
United Kingdom
and Co-Chair:
State of the Science and Technology
Over the last 20 or so years, CAD systems have evolved
from tools for computer aided drafting and the design of
three-dimensional objects to software systems applied to
a wide range of different tasks such as design, planning,
manufacturing, testing, quality assurance and facility man-
agement - i.e. CAx techniques such as CAP, CAM, CAT,
CAQ .... Basically, CAD systems deal with the generation,
manipulation, storage and visualisation of 3D data. They
serve also as database for the integration of various data
acquisition techniques, including photogrammetry. The
interrelation between CAD/CAM and photogrammetry is
bi-directional: on the one side, the results of photogram-
metric processing are used for 3D object modelling in CAD
systems, e.g. for the as-built documentation in architec-
ture and industry, for indoor scene modelling, for visualisa-
tion and animation. In this case, photogrammetric meas-
urement tools are interfaced with, or integrated in, existing
CAD systems. On the other side, CAD modelling concepts
support the (automated or semi-automated) photogram-
metric recognition and measurement process by providing
a priori knowledge of the object.
CAD-based photogrammetry may be defined as a close
combination of CAD and photogrammetry aimed at the
generation of 3D models containing geometry and topol-
ogy. A variety of CAD-based photogrammetry systems
exist offering different measurement and modelling con-
cepts. Efficiency improvement by automation is often lim-
ited in close-range photogrammetry due to the complex
structure of objects. In such cases, semi-automated
approaches are advantageous: the human operator is
responsible for scene interpretation and CAD model
recognition. He selects approximations of objects from the
CAD database and ensures initial fits of the projected
model and the imaged object. The precise object meas-
urement is then carried out by an automatic matching
process.
The increasing interest in generating 3D models has driven
the development of low-cost and user-friendly software
Systems for image measurement and object modelling.
International Archives of Photogrammerty and Remote Sensing. Vol. XXXIII, Part A. Amsterdam 2000. ——
ISPRS
2000
These systems include photogrammetric knowledge as a
"black box" and have been used by non-photogram-
metrists, above all. In this way, the integration of pho-
togrammetric tools with modelling techniques may lead to
a closer partnering between the measurement and the end-
user communities.
References
Chapman, D., Peipe, J., Robson, S., 1999. On the integra-
tion of digital photogrammetry with Computer Aided Engi-
neering. Int. Arch. Photogrammetry & Rem. Sensing, Thes-
saloniki, Greece, Vol. XXXII, Part 5W11, pp. 95-102.
van den Heuvel, FA., 2000. Trends in CAD-based pho-
togrammetric measurement. Int. Arch. Photogrammetry &
Rem. Sensing, Amsterdam, The Netherlands, Vol. XXXIII
Patias, P., Peipe, J., 2000. Photogrammetry and CAD/CAM
in culture and industry - An ever-changing paradigm. Int.
Arch. Photogrammetry & Rem. Sensing, Amsterdam, The
Netherlands, Vol. XXXIII.
WG V/3 ‘Scene Modelling for Visualisation and Virtual
Reality'
by Chairperson: Sabry El-Hakim, National
Research Council, Canada
Wolfgang Fórstner, University
Bonn, Germany
and Co-Chair:
State of Science and Technology
Virtual reality (VR) is enjoying tremendous interest from
researchers and developers across several disciplines,
including photogrammetrists. The applications include:
virtual museums, historical sites documentation, map-
ping of hazardous sites and underground tunnels (mine
automation), as-built modelling of industrial and power
plants for design verification and modification and virtual
production for the entertainment industry. The problems
include practical and precise sensor calibration, the
automatic registration of images, the complete and accu-
rate coverage of all details (which may require multiple
sensors), the automation of the modelling procedure and
maintaining realism during real-time display. Environment
or site modelling is proving to be more difficult than
object modelling because of the size and complexity.
Given the application requirements, selecting and imple-
menting the most efficient method for data collection and
modelling is not obvious, since several methods and a
variety of sensors exist. These vary significantly in the
ability to capture details, cost, accuracy, speed and ease
of use. Selecting the most suitable and efficient method,
along with its configuration, is not easy because for some
of the paradigms actual capabilities and limitations are
not known.
Over the past few years many advances have been
made, both on the scientific and commercial sides. Many
efforts are being made in automating feature extraction,
matching and modelling procedures. However, success
has been limited to image sequences, such as those
obtained by video cameras, and simple scenes and
objects. The model-building procedure still requires a
human in the loop, particularly for sites and complex
objects. Efforts in automating this procedure will rely
mainly on advanced computer vision and scene under-
standing techniques. Commercially, low cost photogram-
—191
Ee NEE ee ——————————Á Mee
