Object classification and image interpretation with an object-oriented system

The authors propose and discuss an innovative integrated architecture of an environment for a multi-user image processing system, based on the extensive usage of the object-oriented paradigm concepts, applied both in image processing programs and object representation in the database. This will result in a tighter correspondence between the abstraction mechanism present in the program and the permanent storage and organization of abstractions into the database, enhancing both performance and safety. In addition the adoption of an object-oriented DBMS will allow for reusability of data models and model sharing for different applications; concurrence of different tasks which is a key point in large and multi-user image processing systems is allowed