Japanese / English

Detail of Publication

Text Language Japanese
Authors Seiichi Uchida, Megumi Sakai, Masakazu Iwamura, Shinichiro Omachi, and Koichi Kise
Title FSA-Guided Optimal Segmentation and Its Application to Camera-Based Character Recognition
Journal Trans. IEICE
Vol. J90-D
No. 8
Pages pp.1966-1976
Reviewed or not Reviewed
Month & Year August 2007
Abstract This paper describes a novel segmentation algorithm for one-dimensional signals. The proposed segmentation algorithm employs an optimization strategy based on dynamic programming (DP). In addition, a finite state automaton (FSA) is introduced for incorporating a priori knowledge of the target signal into the optimal segmentation. Specifically, a state of the FSA corresponds to a segment having a certain property. For example, we can use an FSA with three states, each of which corresponds to a segment whose average signal level is high or middle or low. The relations between adjacent segments (such as the relation that a high-level segment is not adjacent to a low-level segment but to a middle-level segment) are regulated by state transitions. The proposed segmentation algorithm has been applied to a camera-based character recognition task, where each character pattern is printed with stripes whose widths represent the category of the character. The proposed algorithm could successfully detect the boundary of the stripes and provide a recognition rate over 99%.
URL http://search.ieice.org/bin/summary.php?id=j90-d_8_1966
Back to list