In this paper we present a new technique to extract layers in a video sequence. To this end, we assume that the observed scene is composed of several transparent layers, that their...
Deformations in handwritten characters have classdependent tendencies. For example, characters of class “A” are often deformed by global slant transformation and never deforme...
Seiichi Uchida, Mohammad Asad Ronee, Hiroaki Sakoe
In the context of an experimental virtual-reality surgical planning software platform, we propose a fully self-assessed adaptive region growing segmentation algorithm. Our method s...
Global vision systems as found in the small size league are prohibited in the middle size league. This paper presents methods for creating a global view of the world by cooperative...
Markus Dietl, Jens-Steffen Gutmann, Bernhard Nebel
In this paper, we propose a novel technique for video shot segmentation and classification based on the Singular Value Decomposition (SVD). For the input video sequence, we create...