March 19, 2008

This tool offers a user friendly graphical interface for allowing the user to manually annotate both image and video files, using descriptors that are formalized as ontologies. It supports the importing of descriptors from predefined OWL ontology files and enabling the definition and employment of user defined descriptors for both image and video annotation. Furthermore, it provides a user friendly interface for the accurate and undemanding annotation of image, as well as for the live and "frame by frame" annotation of video. As regards to image annotation, this tool enables not only to annotate the whole image but also its image segments with concepts included in a domain ontology. Region merging can be applied in the case where segmentation divides a meaningful object to more than one region. As regards to the annotation of video, all features are provided through three different workspaces opening as different tabs allowing for the annotation of the video as a single entity, the annotation of specific shots and the annotation of specific regions. Special attention has been given to the development of the tool’s input and output functionality, so as to properly serve the BOEMIE specific needs.