A Novel Geometric Information Retrieval Tool for Images of Geometric Diagrams

The description of geometric knowledge in geometric documents often uses a combination of text and diagrams. Compared to the text describing knowledge in natural language or mathematical language, the diagrams may contain even more geometric information, and the recognition of them is a challenge. This paper mainly studies how to automatically extract planar geometric information from images of geometric diagrams. We firstly explore the classification of usual images of diagrams, secondly describe a general framework for geometric information retrieval from images of diagrams, then design and implement a geometric information retrieval tool—GIRTOOLS based on the framework. Finally, experiments demonstrate the feasibility and effectiveness of the approaches implemented in GIRTOOLS. This work may have potential applications in geometry education and knowledge discovery.