Research topics in pictorial databases

Abstract

Pictures are natural and effective means of communication among people, computers, and robotics. A pictorial database is a collection of sharable pictorial data encoded in various formats. During the past several years, pictorial databases have attracted growing attention as an important component in building pictorial information systems as well as intelligent information systems. Eight research tasks are presented. They are comparing the numerical and the linguistic variable approaches, examining new linguistic hedges, studying the similarity of various similarity measures, investigating pictorial data compression techniques, performing pictorial data compression using array grammars, applying the entity‐relationship (ER) model to picture description; further investigating the relationship hierarchy for picture representation using ER diagrams, and extracting pictorial knowledge from pictorial databases. The research results may have a major impact on the development of object‐oriented pictorial databases.