Main Research Areas, Organizational Structure:
The Image Science Division works on the computer-based interpretation
of 2D and 3D image data sets from conventional and non-conventional
image sources. We treat the complete cycle from signals to their
interpretation, and to the resulting action. It is our objective to
develop universal concepts and methods, and it is our strategy to let
difficult applications drive our research and development.
Applications are the touchstone for the true value of a theory and
very often the source of unexpected results wich, in turn, frequently
open up new research directions. A new aspect of our work is the
modeling, simulation and visualization of the underlying scene and
processes; a domain where Computer Vision meets Scientific
Visualization.
The following description of the research projects is structured
according to the working groups of the Image Science Division:
Medical Image Analysis and Visualization
Portal Imaging - a clinical research project funded by the cancer
research of the canton Zürich to improve quality assurance in
radiotherapy treatment. Electronic Portal Imaging Devices (EPID)
enable us to register megavoltage X-ray images of the treatment field
during irradiation. These portal images are then analyzed using a high
precision and area-based matching algorithm in order to measure
patient setup deviations.
PET-MRI
PET-MRI. Further information is available via email.
BIOMORPH - a EU BIOMED-2 Project: Development and Validation of
Techniques for Brain Morphometry, a cooperative effort of 6 European
academic and clincal partners.
Lesion Evolution in Multiple Sclerosis
Lesion Evolution in Multiple Sclerosis - an ETH-KBS project:
The characterization of MS lesion development is mostly based purely
on the spatial pattern of lesions and its variation within the measured
time period. In most cases even only very simple characteristics of
this pattern (as e.g. the total lesion load) is used to find correlation
between clinical and radiological findings.
Lesion development is a complex spatio-temporal process, consequently
methods, concentrating exclusively on the spatial or temporal aspects
of it cannot be expected to provide optimal results.
The major goal of our work is to develop adequate mathematical
frameworks
for integrated spatio-temporal lesion models and to test their power
for better explanation of clinical and immunological findings.
Further information is available via email.
Prostesis Migration
Prostesis Migration.
Further information is available via email.
Interactive Segmentation
Interactive Segmentation:
In spite of considerable efforts during the past decades, image
segmentation
is still one of the major bottlenecks in medical image analysis. Neither
purely manual nor fully automatic approaches are appropriate for the
correct, efficient and reproducible identification of organs in 3D
radiological data volumes. Goal of this pilot project is the exploration
of the power of new man-machine interfacing paradigms as offered
by virtual reality (including graphics, audio and haptics), e.g.,
resulting in
new closed-loop segmentation systems, allowing an optimal cooperation
between computer-based image analysis algorithms and human operators.
Further information is available via email.
Back to the BIWI
Home Page
Shape Modeling and Visualization
LaSSo (LAparoscopic Surgery SimulatOr):
The basic idea of laparoscopic surgery is to minimize damage to healthy
tissue while reaching the actual surgical location. This results in major
gain in patient recovery after operation. The price for this advantage is
paid by the surgeon, who loses direct contact with the operation site.
The operations are usually performed under mono-scopic vision and highly
restricted manipulative freedom, which requires very special skills
from the surgeon. Up to now no appropriate training devices are available,
which would allow to fully acquire these skills before actual intervention
on patients. The goal of the project is the development of a
laparoscopic surgery simulator device using the techniques of virtual
reality which provides nearly realistic training environment.
Bone Morphometry
Bone Morphometry. Further information is available via email.
3D-Skeletons
3D-Skeletons: Skeletonization is an appealing way to describe
and characterize object shape. While computer generation of 2D skeletons is
basically solved, the calculation of skeletons of large discrete 3D objects poses serious
theoretical and technical difficulties. The goal of the project is the
development of procedures for the generation of skeletons of large, complex
natural (mostly anatomical) objects using computational geometrical techniques.
The generated skeletons can then be used e.g. for recognition of organs in 3D
radiological image data. The analysis of the sulcogyral structure of the
human cerebral cortex and brain atlas matching are used as prototypical
application examples. Further information is available via email.
CONQUEST - (Clinical Oncology Network for Quality in European
Standards of Treatment) - a joint European
project with several academic and clinical partners to develop generic,
multimedia-based, quality assurance systems which can be rapidly disseminated
by telematics to all relevant treatment sites so that variations in cancer
treatment outcome can be reduced. The swiss partners will provide anatomical
knowledge representation tools for CT data segmentation in treatment planning
for prostate cancer irradiation.
Back to the BIWI Home Page
Image Understanding
AMOBE - (Automation of Digital Terrain Model Generation and Man-Made
Object Extraction from Aerial Images) - a 3-year ETH project in collaboration
with the Institute for Geodesy and Photogrammetry. The purpose of the
project is to develop semi-automatic and automatic methods for extracting
quantitative 3D information on man-made objects from aerial images.
MINORA - (MINiaturized Optical Range
camera for safety, surveillance and automotive Applications) - a
4-year project in the OPTIQUE II programme of the ETH Council in
collaboration with several academic and industrial partners. The
purpose of the project is to develop range cameras based on optical
ranging techniques (time-of-flight or AM laser radar), with which a
large part of today's safety, surveillance and automotive applications
can be solved.
COGNIS - (COmputer Guided Nannofossil Identification System) - a 3-year
ETH project in collaboration with the Institute of Geology. The purpose
of the project is to develop time-efficient semi-automatic and automatic
methods to find, measure, identify and count micrometer-sized coccoliths
in Scanning Electron Microscope (SEM) images from ocean sediment
specimens.
Back to the BIWI Home Page
Feature and Model Extraction
Image Indexing - an ETH Project: The project "an integrated image
analysis and retrieval system" is a joint project with the computer
systems and the database group of the Computer Science Department
of the ETH. The vision group studies which visual features can
support efficient indexing into an image database. The development
of the appropriate feature extraction algorithms is examined, as
well.
Back to the BIWI Home Page
Remote Sensing
RSIA - (Remote Sensing Image Archive) - an ETH Project: The goal of the project
is to intensify research activities in the
field of content-based picture retrieval from remote sensing image
databases. We investigate the use of
general, robust image features to be used as indexes in a relational
database. In contrast to similar approaches, these features do not
characterize application specific aspects of the images but aim at a
global, signal-oriented description of the data. We intend to perform
the process of information extraction in a full Bayesian approach.
The data handling strategy involves
multiresolution signal representation based on the wavelet transform.
This choice enables efficient
image transmission techniques and supports the signal-oriented query
methods.
New Methods for Runoff Modeling - a Swiss National Foundation Project: the goal of the project is to simulate the effect of a possible climate warming on the areal extent of the seasonal snowcover, the glacier retreat resp. progress and on the runoff regime in the Swiss alps. The concept of the project requires, that GIS as well as remote sensing techniques are involved in combination with the runoff model SRM-ETH.
CARTESIAN - (Cost effective Application of Remote sensing to
enviromenTal aspects of ski rEgions; a Ski region monItoring and
mAnagement informatioN system) - a CEO Project: this project aims
at developing a methodology based on satellite data to provide a
cost-effective assistance in the monitoring and
sustainable maintenance of ski-regions. This includes
issues as: Environmental aspects such as vegetation
indices, landscape changes, snow cover, socio-
economic
issues such as tourism potential and economic
development.
Model Based Snow
Segmentation
Model Based Remote Sensing Image Analysis for Snow Segmentation in
Mountainous Areas - a Swiss National Foundation Project:
Snow segmentation in remote sensing images of mountainous areas is not a
trivial problem. Shadows, reflections and atmospheric effects make the
differentiation of snow, ice and ground very difficult, especially in
regions with rough relieves. The present project aims to develop a new
model driven method to perform a highly accurate segmentation of snow in
optical and SAR remote sensing images. Probabilistic models and estimation
theory concepts represent the basics of this project. Special attention
will be given also to information fusion from these different sensors
and/or channels. Further information is available via email.
Back to the BIWI Home Page