Correspondence problems are one of the key problems in computer vision.
They appear for example in motion analysis, image matching, or stereo vision.
The goal is to assign structures in one image to structures in a second image,
in a way that corresponding structures are as similar as possible.
For this, some property of the pixels has to be assumed to stay
In general the solutions of correspondence problems are nonunique. As everybody can imagine there are several possibilities to match structures. To deal with this problem additional smoothness assumptions have to be applied, which can considerably affect the outcome. For more information see our research page.
The goal of the seminar is to give an introduction to actual solutions of correspondence problems in the areas of motion analysis, image matching and stereo vision. To this end selected work from the literature will be presented and discussed.
||Title of the
||Determing Optical Flow
||An iterative image registration
with an applicaton to stereo vision
||Lukas/Kanade meets Horn/Schunck:
Combining local and global optical flow methods
||Nonlinear matrix diffusion for
optic flow estimation
||An Investiagation of Smoothness
Estimation of Displacement Vector Fields from Image Sequences
||Numerical Methods for Image
Principle Axed-based Registration
||Image Matching as a diffusion
an analogy with Maxwell's demons
||Multi-Modal Volume Registration by Maximization of Mutual Information|
||A theoratical framework for
convex regularizers in PDE-based
computation of image motion
||Performance of optical flow