{"updated":"2025-01-22T05:33:53.465633+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00055161","sets":["1164:4961:4994:4996"]},"path":["4996"],"owner":"1","recid":"55161","title":["古文書OCRのための文字切り出し"],"pubdate":{"attribute_name":"公開日","attribute_value":"2002-07-26"},"_buckets":{"deposit":"0caf43f1-15d3-4144-ba4e-c471f0ba8ac0"},"_deposit":{"id":"55161","pid":{"type":"depid","value":"55161","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"古文書OCRのための文字切り出し","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"古文書OCRのための文字切り出し"},{"subitem_title":"Segmentation of Cursive Characters for Classical Literal OCR","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"2002-07-26","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"国文学研究資料館研究情報部"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"National Institute of Japanese Literature","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/55161/files/IPSJ-CH02055008.pdf"},"date":[{"dateType":"Available","dateValue":"2004-07-26"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-CH02055008.pdf","filesize":[{"value":"308.6 kB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"24"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"38f0bee2-f8c6-4465-bfa3-1d01adbf4cb8","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2002 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"原, 正一郎"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Shoichiro, Hara","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN1010060X","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"本稿では、古文書OCRの前処理として必須である、つづき文字を切り出す手法について提案する。提案する方法では、まず簡単なフィルタ処理（色に基づき文字の候補となるピクセルを抽出するカラーフィルタ、ゴマ塩雑音などの除去フィルタ、カラー画像を白黒階調さらに２値化するフィルタ）により、雑音の少ない良好な２値画像を作成する。次に周辺分布からページごとの平均文字サイズ、縦書き・横書きに関するレイアウト情報を抽出する。文字はこれらの情報に基づいてピクセルから組み立てる。つまり隣接するピクセルを集めて文字のセグメントを生成し、次いで近傍のセグメントを集めて文字あるいはつづき文字を生成する。つづき文字の切り出しは文字輪郭上の相対する凹部分を結ぶ線に沿って行う。本法の特徴は、適切な凹部分を画像の多重解像度解析に基づいて発見するところにある。","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"This paper is a proposal for new means of separating a cursive character into separate and distinctive characters for further OCR processing. The proposed method begins with some filtering, i.e., a color filter to extract candidate pixels of characters according to their color, a noise reduction filter, a conversion of a color image to a gray image, and a binarization. Layout information as to whether a text is written vertically or horizontally as well as average character size (ACS) in a page is obtained from the analysis of a peripherally projected histogram. A character is constructed gradually from pixels. First, connected pixels are aggregated to a small segment. Then neighboring segments are collected to a character or a cursive string. At last, segmentation of a cursive string is basically made along the line connecting the concavity on a contour and its vicinity concavity on the opposite contour. The strength of the new method is to find appropriate concavity by multiresolution analysis.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"56","bibliographic_titles":[{"bibliographic_title":"情報処理学会研究報告人文科学とコンピュータ（CH）"}],"bibliographicPageStart":"51","bibliographicIssueDates":{"bibliographicIssueDate":"2002-07-26","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"73(2002-CH-055)","bibliographicVolumeNumber":"2002"}]},"relation_version_is_last":true,"weko_creator_id":"1"},"created":"2025-01-18T23:18:52.101314+00:00","id":55161,"links":{}}