Rich Annotation of Images

Objective is to significantly enhance the ability to describe, catalog, search and retrieve images through the use of rich annotations (RA) of features in an image. Traditional approaches rely on text-based notes and keywords in metadata; proposed approach will use annotation layer (AL). AL contains drawings and icons to mark features, and domain specific properties to describe the features. Properties of features are described using a structured format with controlled vocabulary. AL is created in a web browser-based Rich Annotation Workbench (RAW) using Scalable Vector Graphics. RAW will foster collaboration between earth scientists and educators, and foster productive use by novice users through ability to: a) overlay multiple AL on image, b) share, reuse and extend domain specific annotation palette.RAW has commercial applications in: Digital Asset Management, telemedicine, and e-learning.Approach is to: create domain specific annotation palettes; enhance our existing iAnnotate platform to create RAW; define RDF-based metadata model for RA; integrate with existing Earth Sciences digital library metadata repositories of NASA-GCMD and DLESE.Anticipated results: Phase I, prove feasibility that RA of features can be metadata for images; Phase II, enhance and commercialize RAW to be an annotation service in a variety of domains.