|
|
|
|
Currently working on my PhD thesis "A
Grand Challenge: Large-scale event
recognition and tracking". Current
research topics include multivariate
statistics for (multi set) decomposition
methods, image deconvolution, object
tracking methods, machine learning for
classification and model selection.
The main applications are
- Extreme weather classification
and forecasting. Time series of
weather radar data and multispectral
satellite imagery, for recognizing
and forecasting heavy precipitation.
- Stemcell tracking. Time series
of images of NPG stemcells captured
using phase contrast microscopy.
Exemplary code for
Matlab
will be
published here when possible.
Publications
Academic dissertations
- Master's thesis.
Vestergaard, Jacob S. (2011),
Improved nowcasting of heavy
precipitation using satellite and
weather radar data, supervisors:
Nielsen, Allan A. (DTU Space), Larsen,
Rasmus (DTU Informatics) and Bøvith,
Thomas (DMI), August 2011, Technical
University of Denmark.
-
Bachelor's thesis.
Vestergaard, Jacob S. (2009),
Subspace Projections of Climate Related
Geodata, supervisors: Nielsen, Allan
A. (DTU Space) and Andersen, Ole B. (DTU
Space), June 2009, Technical University
of Denmark.
Journal articles
-
Vestergaard, Jacob S.; Nielsen, Allan A.
(2012),
Automated invariant
alignment to improve canonical variates
in image fusion of satellite and weather
radar data, Journal of Applied
Meteorology and Climatology, 2012,
[accepted, early online release available].
Conference contributions
-
Vestergaard, Jacob S.; Dahl, Anders L.; Holm, Peter and Larsen, Rasmus
(2013),
Dynamically constrained pipeline for tracking neural progenitor cells,
Conference on Digital Pathology, SPIE Medical Imaging 2013
-
Vestergaard, Jacob S.; Dahl, Anders L.; Holm, Peter and Larsen, Rasmus
(2012),
Pipeline for tracking neural progenitor cells,
Workshop on Medical Computer Vision, MICCAI 2012
-
Vestergaard, Jacob S.;
Dahl, Anders L.;
Larsen, Rasmus and
Nielsen, Allan A. (2012),
Classification
of Polarimetric SAR Data Using
Dictionary Learning.
In
Proceedings of SPIE, the International
Society for Optical Engineering
Vol. 8537, 2012, p. 8537-35
-
Nielsen, Allan A.; Vestergaard, Jacob S.
(2012),
Parameter optimization in the
regularized kernel minimum noise
fraction transformation, IEEE
International Geoscience and Remote
Sensing Symposium, IGARSS 2012, Munich,
Germany, 22-27 July 2012. Invited
contribution.
- Nielsen, Allan A.; Larsen, Rasmus
and Vestergaard, Jacob S. (2011),
Sparse principal component analysis in
hyperspectral change detection,
presented at SPIE Europe Remote Sensing
Conference 8180, September 2011, Prague,
Czech Republic.
- Nielsen, Allan A.; Vestergaard,
Jacob S. and Andersen, Simon G. (2010),
Bidrag til automatisering af
ortofotomosaikfremstilling,
presented at Kortdage 2010, Århus,
Denmark.
Posters
- Vestergaard, Jacob S.; Dahl, Anders
L.; Holm, Peter and Larsen, Rasmus
(2012),
Large scale tracking of stem cells using
sparse coding and coupled graphs,
presented at International Computer
Vision Summer School (ICVSS), July 2012,
Sicily.
- Vestergaard, Jacob S. (2011),
Improved
Nowcasting of Heavy Precipitation Using Satellite
and Weather Radar Data, supplementary poster for
"Månedens grønne projekt", October 2011, Technical
University of Denmark.
- Vestergaard, Jacob S.; Nielsen, Allan A.;
Larsen, Rasmus and Bøvith, Thomas (2011),
Improved Nowcasting of Heavy Precipitation Using
Satellite and Weather Radar Data, part of
master's thesis, presented at Industrial Visionday,
May 2011, Technical University of Denmark.
- Vestergaard, Jacob S. and Andersen, Simon G.
(2010),
Normalization of ortophotos based on no-change
pixels using Multivariate Alteration Detection
(MAD), presented at Industrial Visionday, May
2010, Technical University of Denmark.
- Vestergaard, Jacob S.; Nielsen, Allan A., and
Andersen Ole B. (2010),
Seventeen years
of global SSH anomalies analyzed by a maxmaximum
information based extension to EOF analysis,
presented at 2010 Ocean Sciences Meeting, February
2010, Portland, OR.
Software
-
hypersph2cart.m and
cart2hypersph.m for
Matlab.
Convert hyperspherical coordinates to
cartesian coordinates and convert
cartesian coordinates to hyperspherical
coordinates.
- The
CLOUD toolbox for
Matlab.
Methods for reading Meteosat Second
Generation Level 1.5 HDF5 (.h5) images
and using included metadata to perform
projections using the
PROJ.4 library. Minimum version 4.4
of the library is needed, as
MATLAB
does not
support the geostationary satellite
projection (GEOS).
Exemplary
weather radar data and multispectral
satellite data is included (thanks to
DMI).
Teaching
2012
2011
|
|
|
|
|
|
|
|
|
|
