Abstract
We explore a simple statistical method for classifying handwritten numerals. We work with numerals only in segmented form. We defined a pseudodistance function between handwritten numerals and classify each numeral based on its distance from the those in the reference set. This method, and its obvious variants, are studied on the basis of their performance on a data-set of moderate size (95 observations). The results suggest that this method may be considered in recognition of handwritten numeric characters.
Get full access to this article
View all access options for this article.