This paper describes an original approach to the problem of edge-based binocular stereovision. The tokens to be matched are subchains of the chains of connected pixels. Local constraints of the stereovision problem are first used in associating to each token a set of potential matches. Global constraints are embedded in a cost function and we look for the minimum of this cost function. The optimisation search is conducted using genetic algorithms.
CITATION STYLE
Vaillant, R., & Gueguen, L. (1994). Genetic algorithms applied to binocular stereovision. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 801 LNCS, pp. 193–198). Springer Verlag. https://doi.org/10.1007/bfb0028351
Mendeley helps you to discover research relevant for your work.