| 大会名称 |
|---|
| 2010年 情報科学技術フォーラム(FIT) |
| 大会コ－ド |
| F |
| 開催年 |
| 2010 |
| 発行日 |
| 2010/8/20 |
| セッション番号 |
| 1L |
| セッション名 |
| 文字認識と画像照合 |
| 講演日 |
| 2010/09/07 |
| 講演場所(会議室等) |
| L会場（総合学習プラザ2F 第15講義室） |
| 講演番号 |
| H-010 |
| タイトル |
| Preliminary Experiment on Khmer OCR |
| 著者名 |
| KRUY Vanna, KAMEYAMA Wataru, |
| キーワード |
| Khmer, OCR, Segmentation, Pattern Recognition |
| 抄録 |
| OCR technology has already matured for major languages such as Japanese and English. However, there is no reliable OCR system for Khmer Language. This is largely due to the lack of Khmer OCR research efforts and the complex nature of Khmer characters. Segmentation is obviously a big issue. In this study, we tackle this issue using Connected Component Analysis and Word Semantic. From text corpus, connected components of characters are extracted and annotated to form Component Dictionary and Word Semantic Database. Preliminary experiment is designed and conducted. In this paper, we present our early results, considerations, and future work. |
| 本文pdf |
| PDF download (540.5KB) |