Randomized and Deterministic Approaches to The Sparse Correspondence Problem

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F07%3A03138374" target="_blank" >RIV/68407700:21230/07:03138374 - isvavai.cz</a>

Result on the web

—

DOI - Digital Object Identifier

—

Result language

angličtina

Original language name

Randomized and Deterministic Approaches to The Sparse Correspondence Problem

Original language description

We summarize recent progress in the image features correspondence problem and we present a robust solution based on a radically different notion than a standard optimization or robust statistics methods. Even if the method represents measurement uncertainty explicitly, the existence of a unique solution is guaranteed. We formulate the correspondence problem as a graph theory problem of a finding stable independent vertex set (strict sub-kernel) in an oriented graph. The vertices of the graph are all possible correspondences, the edges capture the structure of the constraints and edge orientation represents pairwise comparison ''is better'' based on correspondence quality, including the uncertainty of this comparison. Our method allows encoding both non-parametric and parametric constraints with unknown parameters to the graph.

Czech name

Randomized and Deterministic Approaches to The Sparse Correspondence Problem

Czech description

We summarize recent progress in the image features correspondence problem and we present a robust solution based on a radically different notion than a standard optimization or robust statistics methods. Even if the method represents measurement uncertainty explicitly, the existence of a unique solution is guaranteed. We formulate the correspondence problem as a graph theory problem of a finding stable independent vertex set (strict sub-kernel) in an oriented graph. The vertices of the graph are all possible correspondences, the edges capture the structure of the constraints and edge orientation represents pairwise comparison ''is better'' based on correspondence quality, including the uncertainty of this comparison. Our method allows encoding both non-parametric and parametric constraints with unknown parameters to the graph.

Type

O - Miscellaneous

CEP classification

JD - Use of computers, robotics and its application

OECD FORD branch

—

Project

—

Continuities

V - Vyzkumna aktivita podporovana z jinych verejnych zdroju

Publication year

2007

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů