Abstract
The tundamental problem of modelling data effec tively in the context of still pictures is addressed, which marks a significant departure from the conventional means of imple menting pictorial databases. In using the entity-attribute-rela tionship model, the main semantic concepts of entity. at tribute and relationship can be represented in a simple man ner which correspond very closely to the noun, adjective and verb which are the essential components of a simple descrip tion of a picture. Such a system has the advantage over traditional ways of representing the data in an automated system in that it allows, inter aha, the concept of relationships between two (or more) objects to be successfully represented in the database. Our basic starting point is man-machine cooperation in which the pattern processing capability of the human eye is fully exploited As a result, it should provide a much better degree of flexibility in the identification of com plex patterns. Evaluation experiments carried out on a proto type system based on this approach are able to yield precision and recall performances of over 62% and 88% respectively.
Get full access to this article
View all access options for this article.