PENS Repository

A Semantically-Related Information-Extraction System of Living

Mori, Masahiko and Sasaki, Shiori and Kiyoki, Yasushi (2011) A Semantically-Related Information-Extraction System of Living. In: IES 2011-Emerging Technology for Better Human Life, EEPIS-Surabaya.

[img]
Preview
PDF - Published Version
Download (9Mb) | Preview

    Abstract

    This paper presents a semantically related information-extraction system of living things by the global analysis of spatial, temporal and color information of images. The progress of multimedia, mobile an GIS technology makes it possible to and share various information resources globally. Various kinds of information resources on natural environments of the real world are also existing in a cyber space, and it is becoming possible to support users to acquire the valuable knowledge that bridge user’s fragmentary information about the real world and adequate information on the cyber space. This system realizes the functions for identifying unknown living things contained in a picture image input by a user, through the global analysis of temporal, spatial and color information of this images within a user-selected domain, Given a picture image of a living thing with temporal and spatial information, this system evaluates possible candidates of living things. This system also analyzes color information by calculating correlations between the color distribution of an input image and corresponding sample image data. By these processes, users acquire detailed information such as the name, the habitant and the active period of the living things contained within the given images.

    Item Type: Conference or Workshop Item (Paper)
    Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
    Q Science > QA Mathematics > QA76 Computer software
    Divisions: Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science
    Depositing User: Satria Hardinata
    Date Deposited: 14 Nov 2011 12:23
    Last Modified: 14 Nov 2011 12:23
    URI: http://repo.pens.ac.id/id/eprint/1443

    Actions (login required)

    View Item