Membership function adaptive fuzzy filter for image sequence stabilization

doi:10.1109/TCE.2004.1277834

Membership function adaptive fuzzy filter for image sequence stabilization

2004-02-01

29

PRE

AI

DOI：10.1109/TCE.2004.1277834

Original

Original request for help

Save

Abstract

En 中文

In this paper a novel adaptive adaptive fuzzy image sequence stabilization system is proposed for the motion correction part of digital image stabilizers. The proposed stabilizer is based on smoothing of absolute frame positions. Initially a short mean filter is applied to raw absolute frame displacements as pre-process, to reduce the dynamic range of the fuzzy system input. This enables the fuzzy system to give appropriate responses with a fewer number of membership functions, hence provides improved performance at reduced computational load. Fuzzy stabilization is then achieved through fuzzy correction mapping. Output membership functions of the fuzzy system are continuously adapted so as to constitute a membership function adaptive fuzzy filtering process. The proposed membership function adaptive fuzzy filter based image sequence stabilization system is shown to provide excellent stabilization as well as intentional camera movement preservation performance, superior to previously reported systems.(1).

Keywords:

fuzzy logic

image sequence analysis

nonlinear estimation

nonlinear filters

Membership function adaptive fuzzy filter for image sequence stabilization

Abstract

Journal

Organization