RT Conference Proceedings
T1 Reducing the amount of input data in traffic sign classification
A1 Granados, Ana
A1 Ledezma Espino, Agapito Ismael
A1 Gutiérrez Sánchez, Germán
A1 Sanchis de Miguel, María Araceli
AB Several complex problems have to be solved in order to build Intelligent Transport Systems. Among them, it is worth mentioning the detection and classification of tra±c signs which could appear at any position within a captured image. This paper analyzes the influence of the number of attributes in the field of classification of tra±c signs when automatic learning techniques are used. In order to face this task, four different approaches have been considered, three of them symbolic and one sub-symbolic. These techniques have been applied using two different input pattern dimensions and their performances have been compared.
YR 2006
FD 2006
LK https://hdl.handle.net/10016/11217
UL https://hdl.handle.net/10016/11217
LA eng
NO Proceeding of: 3th International Conference Modeling Decisions for Artificial Intelligence, MDAI 2006. Tarragona, Catalonia, Spain, april 3-5th, 2006.
NO The research reported here was carried out as a part of the research project CICYT TRA2004-07441-C03-01.
DS e-Archivo
RD 17 jul. 2024