2026-03-11T06:56:26Z https://kait.repo.nii.ac.jp/oai

oai:kait.repo.nii.ac.jp:00001000 2025-06-17T05:36:05Z 2:16:41:120

手書き筆記体からの文字切り出しおよび認識 Hand-written Cursive Character Segmentation and Recognition 小泉, 和也 Koizumi, Kazuya 西村, 広光 Nishimura, Hiromitsu 富川, 武彦 Tomikawa, Takehiko hand-written character segmentation cursive character character recognition application/pdf The purpose of this paper is an improvement of a handwritten cursive character segmentation method for its recognition system. The paper was evaluated segmentation from human decision in the past. Now, this paper is evaluated segmentation from character recognition without the help of human decision. Our segmentation method is an alphabet cursive character recognition system in cooperated with segmentation and evaluated segmentation result. Constructed character recognition system is able to use word slant correction. The same way, it was compared character slant correction. Both recognition performances are compared in order to evaluate from recognition experiments. As a result of the evaluation, segmentation precision has been improved by using character slant correction. 神奈川工科大学 2006-03-20 jpn departmental bulletin paper VoR https://doi.org/10.34411/00000993 http://hdl.handle.net/10368/994 https://kait.repo.nii.ac.jp/records/1000 10.34411/00000993 AN10074179 09161902 神奈川工科大学研究報告.B,理工学編 30 67 72 https://kait.repo.nii.ac.jp/record/1000/files/kkb-030-012.pdf application/pdf 2.6 MB 2010-02-03