An Intelligent Search Engine for Finding Graphics Objects

Abstract

The difficulties of locating a specific graphics object in a library containing files of graphics objects are well known. Current approaches to solving this problem are labor intensive and require the development of textual taxonomies which can be manipulated by existing textual search engines. The intelligent graphics search engine described in this research utilizes Quinlan's C4.5, a well-known machine learning algorithm, to "team" features of a graphics object. Using these features, the search engine finds all occurrences of the graphics object within a file or group of files. Object recognition can be made regardless of the orientation or scale of the objects being compared. The search engine requires files be in a computer-aided design (CAD) format where graphics primitives are organized into graphics objects. The system automatically revises its knowledge as necessary.

Department(s)

Mathematics and Statistics

Document Type

Article - Journal

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2024 The Authors, All rights reserved.

Publication Date

01 Dec 1998

This document is currently not available here.

Share

 
COinS