Title :
A graph grammar to recognize textured symbols
Author :
Sánchez, Gemma ; Lladós, Josep
Author_Institution :
Dept. d´´Inf., Univ. Autonoma de Barcelona, Spain
fDate :
6/23/1905 12:00:00 AM
Abstract :
This paper describes a graph grammar to modelize textured symbols in a graphics recognition framework. A textured symbol means a symbol consisting of repetitive structured patterns. We propose a method to infer a graph grammar from a structured texture detected in a document, and the subsequent parser to decide whether a symbol is accepted by the grammar. The grammar is based on a region adjacency graph representation of the vectorized document and the productions are based on the neighboring relations of the patterns forming the textured symbol. The syntactic framework is applied on an architectural plan understanding application
Keywords :
character recognition; graph grammars; graph theory; inference mechanisms; architectural plan understanding; graph grammar; graphics recognition framework; parser; region adjacency graph; repetitive structured patterns; structured texture; syntactic framework; textured symbol recognition; vectorized document; Computer graphics; Computer vision; Context modeling; Filling; Mars; Pattern recognition; Production; Prototypes; Tiles;
Conference_Titel :
Document Analysis and Recognition, 2001. Proceedings. Sixth International Conference on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7695-1263-1
DOI :
10.1109/ICDAR.2001.953833