Two "Adjunkt" Vacancies
(Assistant Professors)
in Medical Image Analysis and in Computer Vision
(Robot vision or Virtual Reality)

The Faculty of Technology and Science, Aalborg University, Denmark,
hereby announces two vacant "Adjunkt" positions at the Institute of
Electronic Systems to be filled by December 1, 1996, or as soon as
possible there after (positions Nos. 96015 and 96011).

The research and teaching of the institute covers communication and
speech technology, mathematics and computer science, image analysis and
computer vision, medical informatics and engineering, intelligent
multimedia and process control.

Both positions are associated with the Department of Medical
Informatics and Image Analysis (MIBA). The research areas of the two
positions are within general and medical image processing and analysis,
which also covers pattern recognition, computer and robot vision as
well as their applications in the context of multimedia and "virtual
reality". The department is involved in several international research
projects and networks within both computer vision and medicine and has
many close collaborators in both Europe, the USA, and Israel.

The position no. 96015 within Medical Image Analysis.
This position concerns methods for 3-D reconstruction of bodysegments
(static and dynamic) as well as visualisation in connection with
biomechanical measurements. This research forms part of a very close
collaboration between the Laboratory of Image Analysis (LIA) and the
Centre for Sensori-Motor Interaction (SMI), both at MIBA, and it
includes biomechanics as described in the note attached below. Further
information on the medical and biomechanical aspects of the positions'
research areas can be obtained by email from research council professor
Thomas Sinkjaer ( and research ass. prof. Michael Voigt
( or by telephone (int. +45) 98 15 85 22.

The position no. 96011 within Computer Vision.
The research of this position aims at participation in and
contributions to the broader research areas of the Laboratory of Image
Analysis. However, applicants with interest and experience in vision
guided robot navigation and active vision will be preferred.
Alternative priority is given to modelbased vision in application to
intelligent multimedia and "virtual reality". A supplementary note on
these research areas is also attached below.

Teaching (Both positions)
The teaching obligations of these positions will mainly be within the
B.Sc. and M.Sc. programmes of electronic and computer engineering and
with specialisation in image analysis and computer vision, but may also
be within other study programmes of the university.

Further information regarding computer vision as well as general
aspects of both positions may be obtained from professor Erik Granum
( and ass. prof Claus B. Madsen ( by
e-mail or telephone (int. +45) 98 15 85 22.

The adjunkt position (assistant professor) and qualifications required.
In Denmark this is a 3-year-post, through which the candidate could
achieve sufficient teaching experience and establish an independent
research record, such as to obtain the necessary qualifications to
apply for tenured positions.

An appointment as Adjunkt presupposes relevant academic qualifications
at Ph.D., or comparable, level. Teaching experience is desirable but
not required. The total assessment, of each applicant, emphasizes the
research potential. Normally such posts are offered only to applicants
with less than 8 years seniority (from the award of the first academic

General information and conditions:
A copy of the memorandum on the conditions of appointment of academic
staff, with research and teaching obligations, to posts at academic
institutions, may be obtained from the Faculty Office (,
telephone no. (int. +45) 98 15 85 22, Ext. 7471.

Appointment is according to the Agreement on Salaries and Conditions of
Appointment, between the Ministry of Finance and the Central
Organisation of Academic Unions (AC).

Applicants for the posts should define the basis on which they wish to
be assessed, as far as their academic, teaching, and other
qualifications are concerned.

Furthermore, applications should contain documentation of:

- Educational qualifications.
- Academic experience. A list of publications should be included,
in which the applicant indicates what is submitted as documentation
of relevant experience.
The Assessment-Committee may request further documentation.
- Teaching experience, if any.

as well as
- Personal information
- Information on previous employment, etc.

Assessment of the applications is made by an Expert Assessment
Committee. Its recommendation is sent, in full, to all applicants of
the same position.

Applicants, who wish to be considered for both positions, may indicate
this accordingly.

Applications, with reference to the Appointment number(s), and
containing 3 copies of each of the documents and publications, which
are to be included for the assessment, as well as a list of all the
material submitted, should be received at the university, no later than
by the morning mail, on

- October 31, 1996

Applications should be addressed to:

Aalborg University
Faculty of Technology and Science
Postbox 159

*______________*________________*_________________ *___________*

Supplementary Note
Re.: Position No. 96015, Medical Image Analysis

Human motion is a result of the interaction between sensory
information, central motor programming, muscle activation and the
biomechanics of the musculo-skeletal system. Analysis of the complex
control of normal and impaired human movement is conditioned on methods
which combine computerized visualisation techniques and movement
simulations based on anatomical and physiological models of the
musculo-skeletal system.

Description of human motion is today achieved through the use of
dedicated image analysis systems that employ markers placed on the
human body at selected positions. Through monitoring of the motion of
the markers, it is possible to describe the motion in terms of a stick
figure model. The stick figure information is combined with
anthropometric information about some general and some specific human
dimensions to enable computation of joint torques, mechanical work
production etc. However, such models are very coarse approximations to
the individual subjects studied, and in some cases so inaccurate that
meaningful analysis is questionable.

It is the ambition to achieve subject-specific motion models i.e.,
models based on the "true" individual anthropometric information about
segment volumes, positions of mass centers, radii of gyration etc. for
the individual subjects. This assumes methods for automatic acquisition
of volumetric information of the human body. The model can either be
based on stereometric information, magnetic resonance imaging or a
combination of these. Such methods will enable construction of more
accurate subject specific models, and potentially the need for markers
may be eliminated.

Simulations of movements of large scale musculo-skeletal systems, i.e.
the human body, should be based on anatomical and structural
information about functional anatomy of the human musculo-skeletal
system and non-linear dynamic models of muscles, sensors, and
activation processes.

A research project addressing these issues is in progress as a joint
venture between the Center for Sensory-Motor Interaction and the
Laboratory of Image Analysis, both at the Department of Medical
Informatics and Image Analysis. In support of this multi-disciplinary
project the assistant professor position has been allocated. The
research in this position should focus on the extraction of volumetric
information of parts of the human body, real-time estimation and
visualisation of human motion, and movement simulation based on
physiological/biomechanical models.

>From a geometric point of view the human body is a deformable and
articulated object. It is expected that the methods developed for
monitoring human motion, will be applicable also in other application


Supplementary Note
Re.: Position No. 96011, Computer Vision:

The ambition of the research in computer vision is the continually
operating and integrated vision system, which in real time interprets
dynamic scenes. The approach exploits active vision techniques, focus
of attention, and temporal context. The major demonstrator projects of
the Laboratory of Image Analysis are vision guided navigation using a
RoboSoft Robuter20 mobile platform and active cameras. Purposive
approaches have dominated the recent years of research. Active vision
is experimentally investigated with a 10 DoF stereo camerahead. Various
techniques for camera control and view point selection have been
developed. Foveated vision and visual servoing are also on the
research agenda.

New research initiatives at the Laboratory of Image Analysis are
directed towards applications in connection with intelligent
multimedia, virtual reality, and visualisation. Gesture recognition and
interpretation using modelbased approaches are in focus for various
reasons, one of which is the ambition to provide for real time
visualisation of vision generated/updated models of observed gestures
and/or body movements.

Michael Voigt, ph.d.,
ass. research prof.
Center for Sensory-Motor Interaction, Aalborg University
Fredrik Bajersvej 7D, 9220 Aalborg East, Denmark
Tel. : direct + 45 98 15 42 11 and after the ready tone : 4968
operator :+ 45 98 15 85 22, ext. 4968
Fax : + 45 98 15 40 08