Ad: Hyperspectral Postdoc position [message #27865] |
Mon, 05 November 2001 15:53  |
Pete[1]
Messages: 7 Registered: April 1999
|
Junior Member |
|
|
CSIRO Mathematical & Information Sciences
and CSIRO Exploration & Mining
Postdoctoral Fellowship in Hyperspectral Image Analysis
$49 - $55K + Superannuation
The Divisions of Mathematical & Information Sciences (CMIS), and Exploration
&
Mining (CEM) are part of Australia's largest scientific and industrial
research
organisation - CSIRO. As a world leading and diverse scientific research
institution, our emphasis is on bringing teams together from different
scientific fields to generate solutions to problems facing Australian
industry,
and the rest of the world.
We are seeking to appoint a high-calibre Postdoctoral Fellow in the area of
hyperspectral image analysis of very large datasets for a period of 3 years,
starting in 2002. The position is located at North Ryde in Sydney,
Australia.
Airborne and spaceborne hyperspectral images are extremely large,
multi-dimensional data sets with spatial structure. CEM have been using
them
for a number of years to map regions of geological and mining interest.
Their
customers in the global mining sector are now providing them with even
larger
hyperspectral data sets, typically many million pixels in 100-200 spectral
bands. Single flight-line datasets are thus not uncommonly measured in
gigabytes and whole surveys in hundreds of gigabytes. Such huge volumes of
data
have created the need to automate the processing of such data sets and the
creation of information products, and to do so rapidly (ie. in minutes not
hours).
While some initial pioneering work has been carried out in this area, by
CSIRO
scientists and their collaborators, there is a serious need to increase the
effort to keep up with the demands of ever larger data sets now routinely
being
collected by users, and new geological and environmental questions arising
from
them.
Particular topics that the successful scientist will address include:
1. The automated identification of landscape materials (e.g. minerals,
rocks,
vegetation, man-made targets) in a scene to sub-pixel level with and without
the aid of spectral libraries, and with minimal a-priori knowledge.
2. The incorporation of spatial context into the methodologies.
3. The development of hyperspectral data mosaicing and content matching
methods.
The scientist will also be expected to implement prototypes of the more
important methods he/she develops in C or C++, and assist with their
integration into existing packages.
Selection Criteria
Essential
1. A recent Ph.D in a relevant quantitative discipline.
2. Extensive knowledge of modern multivariate statistical methods.
3. Good C programming skills.
Desirable
4. Experience with hyperspectral and/or very large spatial data sets.
5. Experience with packages such as IDL or S-Plus.
For further information about this project, please contact Mark Berman, on
(02) 9325 3205, email: Mark.Berman@csiro.au or Jon Huntington on
(02) 9490 8839, email: Jon.Huntington@csiro.au .
A comprehensive 3-year Development Plan will be provided, ensuring that over
the period of the Fellowship you will develop skills, partnerships and
networks
which are valued both within CSIRO and industry.
Applications must include your CV, with a covering letter outlining skills /
experience relevant to the project, and the names of at least two
professional
referees. Please forward your application, quoting the reference number
01/S46, to jobs@cmis.csiro.au by Tuesday, 13 November, 2001.
|
|
|
Re: Ad: Hyperspectral Postdoc position [message #27895 is a reply to message #27865] |
Fri, 09 November 2001 07:49  |
John-David T. Smith
Messages: 384 Registered: January 2000
|
Senior Member |
|
|
Paul Manusiu wrote:
>
> Yes the world does exists outside the USA, and this world may have their own
> definition of words !
Having spent two glorious weeks down under this summer (your winter), I
recall my favorite Aussie roadsign, encouraging clean roadways
(especially those made of "bitumen"): "Do your bit". I, for one, try to
do my bit everyday.
JD
|
|
|