Two-dimensional rational automata: A bridge unifying one- and two-dimensional language theory

4Citations
Citations of this article
2Readers
Mendeley users who have this article in their library.
Get full text

Abstract

We define two-dimensional rational automata for pictures as an extension of classical finite automata for strings. They are obtained replacing the finite relation computed by the transition function with a rational relation computed by a transducer. The model provides a uniform setting for the most important notions, techniques and results presented in the last decades for recognizable two-dimensional languages, and establishes new connections between one- and two- dimensional language theory. © 2013 Springer-Verlag Berlin Heidelberg.

Cite

CITATION STYLE

APA

Anselmo, M., Giammarresi, D., & Madonia, M. (2013). Two-dimensional rational automata: A bridge unifying one- and two-dimensional language theory. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7741 LNCS, pp. 133–145). https://doi.org/10.1007/978-3-642-35843-2_13

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free