Segmenter for Known Number of Characters
Original Publication Date: 1984-Dec-01
Included in the Prior Art Database: 2005-Feb-05
In the recognition of handwritten symbols, it is necessary to have an efficient procedure for the segmentation of multistroke symbols into isolated symbols. A previously described segmentation procedure [*] used a fixed segmentation threshold to separate the characters of a line of writing purely on a spatial basis. For that procedure, segmentation errors are of two types. On the one hand, a single character can be split into two segments. For example, the B of Fig. 1 might be split into 1 and 3. On the other hand, two characters could be grouped into one segment. For example, C and D in Fig. 1 are somewhat close and might be considered one segment. Experimentally, it has been found that some writers space characters closely and others space them more widely apart.