Modeling of figure-ground information is a key issue in the field of cognitive vision and visionbased object recognition. We have used rough set derived lattices in order to express such figureground relations. Complement relations in lattices turn out to be suited for modeling such relations. We have applied this method to a virtual world created by a literary text. We have investigated the figure-ground property of characters in the mystery novel The Murders in the Rue Morgue by Edgar Allan Poe. We have found that in this story, the characters have a clear figure-ground relation most of the time. This can be explained by the style of writing where the culprit does not blend in with the rest of the character from the beginning of the story. In other words, most of the characters' attribute information is different. Due to our lattice representation, the evolution of the story can be expressed numerically in terms of figure and ground.
|ジャーナル||International Journal of Advancements in Computing Technology|
|出版ステータス||Published - 2012 12 1|
ASJC Scopus subject areas
- コンピュータ サイエンス（全般）