Similarity‐driven pictorial database design using entity‐relationship approach

The entity‐relationship approach is applied to similarity‐driven pictorial database design. The capabilities and the limitations of computers are presented. Advantages of the entity‐relationship approach, and similarity‐driven pictorial database design are also presented. Classification of pictures, similarity‐driven face databases, chromosome databases, leukocyte databases are investigated. Quantitative measure of equal‐perimeter circular shape, equal‐area circular shape, elongated, spiculed, indented, slightly indented, and deeply indented leukocytes are also investigated. The entity‐relationship diagrams of a house, a house with high roof, and a church are shown as illustrative examples of this approach. The class of elongated isosceles triangles is viewed as a “fuzzy entity”. Topics for future research are also stated. The results have useful applications in data engineering, knowledge engineering, pattern recognition, artificial intelligence, fuzzy language theory, computer graphics, expert systems and pictorial database design.