2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 7
Presentation Time: 9:45 AM

RICH ANNOTATION SERVICE USING XIMA


JAIN, Pramod, Innovative Decision Tech. (INDENT), 9000 Cypress Green, Jacksonville, FL 32256, pramod@indent.org

A web-based rich annotation service for images that uses XML for Image and Map Annotation (XIMA, a proposed engineering specification to OGC’s interoperability program) will be presented. Rich annotations are a method to identify and describe the content/features of images. Features are identified by graphical annotations that are drawings or icons placed by user in a transparent layer on an image. Features are described by domain specific properties, whose values may be obtained by web services that provide an interface to datasets.

Rich annotations can be used to a) communicate information in a graphical manner about features and their attributes in an image and b) create content-based metadata for images such that images can be intelligently cataloged, searched and retrieved based on content.

In both uses, XIMA will provide the necessary interoperability such that a) any XIMA compliant server will be able to process the annotation record, and b) annotation can be overlaid on another image or map

This presentation will focus on the capabilities of XIMA and presenter’s experience with using XIMA for: a) the intended purpose of sharing annotations, and b) creating content-based metadata. The basic structure of XIMA will be presented, followed by a detailed walk through of a XIMA document. XIMA is strong in its ability to capture the geometry and location of annotations, and its ability to describe the same content in two more images. It has limitations in its ability to describe a wide variety of attributes that may be associated with an annotated feature.