Lindeberg, ``On the axiomatic foundations of linear scale-space''.
Technical report ISRN KTH NA/P--93/18--SE.
Revised version published as
Chapter 6 in J. Sporring, M. Nielsen, L. Florack, and P. Johansen (eds.)
Gaussian Scale-Space Theory: Proc. PhD School on Scale-Space Theory,
(Copenhagen, Denmark, May 1996), pages 75--98,
Kluwer Academic Publishers, 1997.
(PostScript 132 kb)(PDF 394 kb)
Feature detection, automatic scale selection and scale-invariant image descriptors
Feature detection methods based on the combination of Gaussian derivative
operators at multiple scales. Special focus is given to the problem of
scale selection, in order to adapt the local scales of processing
to the local image structure.
Specifically, the notion of automatic scale selection based on
gamma-normalized derivatives makes it possible to define
scale-invariant image features.
The use of such scale-invariant image features allows the vision system to
automatically handle the unknown scale variations that may occur in real-world
image data, due to objects of different physical size as well as objects with
different distances to the camera.
Laptev and Lindeberg:
``A distance measure and a feature likelihood map concept for scale-invariant model matching'',
International Journal of Computer Vision, vol. 52, number 2/3, pages 97--120, 2003.
(PDF 1.5Mb)
Spatio-temporal analysis and velocity adaptation with application to recognition of activities
Direct methods for recognizing spatio-temporal events with associated activities based on the local spatio-temporal image structure, without explicit inclusion of tracking mechanisms or other temporal trajectories. In order to handle a priori unknown relative motions relative to the observer, a notion of local velocity adaptation is introduced.
Estimation of affine image deformations and direct computation of cues to surface shape (with theory for multi-scale second moment matrices/structure tensors and affine shape adaptation)
Theories and algorithms for
shape from texture and shape from disparity gradients
based on affine distortions of local 2-D brightness structure.
Specifically, this framework includes a theory for local affine normalization
of local image descriptors by affine shape adaptation, which makes
it possible to define affine invariant image features.
These papers also outline the theory for multi-scale second-moment
matrices also referred to as structure tensors.
T. Lindeberg and L. Bretzner,
``Forfarande och anordning for overforing av information
genom rorelsedetektering, samt anvandning av anordningen'',
Swedish patent 9800884-0, March 1998 (now released).
Bretzner,
Laptev, Lindeberg, Lenman and Sundblad:
``A prototype system for computer vision based human computer interaction'',
Technical report CVAP251, ISRN KTH NA/P--01/09--SE. Department of Numerical Analysis
and Computer Science, KTH (Royal Institute of Technology), S-100 44 Stockholm, Sweden,
April 23-25, 2001.
Demo presented at the Swedish IT-fair Connect 2001, Ãlvsjömässan
, Stockholm, Sweden, April 2001.
(PDF 522kb)
Using combinations of scale selection and model-based ridge detection for
blood vessel segmentation (Alejandro Frangi et al at Utrecht University) and
detection of tubular structures (Karl Krissian et al at INRIA Sophia Antipolis) in medical 3-D images.