Abstract
In many image classification problems the extent of use- fulness of any variable for the purposes of discrimination apriori is unknown. This paper describes a unique fuzzy rule generation system developed to overcome this problem. By investigating interclass rela- tionships very compact rule sets are produced with redundant variables removed. This approach to fuzzy system development is applied to two problems. The first is the classification of the Fisher Iris data [4] and the second is a road scene classification problem, based on features ex- tracted from video images taken by a camera mounted in a motor vehicle.
Author supplied keywords
Cite
CITATION STYLE
Wilson, M. (2001). Interclass fuzzy rule generation for road scene recognition from colour images. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2124, pp. 692–699). Springer Verlag. https://doi.org/10.1007/3-540-44692-3_83
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
