OCR and translation for archive collections

Catalogues, historical documents and foreign-language archives can combine image-quality, layout and language challenges. Evaluate a few representative pages before extending the workflow.

This page describes uses of existing Geyi tools. The material owner remains responsible for processing rights, sensitive content and final review.

Start by checking a sample

Choose authorized samples of text, tables and complex layouts, or explore our fixed samples without signing in.
Extract text and compare missing characters and reading order with the source.
Choose preserved-layout or reflow translation; check names, places and terminology.
Keep outputs associated with their originals and obtain the material owner’s review before use.

Before processing your material

Scan quality, historical characters, handwriting and complex layouts can affect results. Evaluate representative pages; OCR does not replace archival review.

Register or sign in to process your own material. Current tools and your account show limits and pricing. Use the contact page to discuss material types; this page makes no commitment to private deployment, institutional pricing or bespoke delivery.

See it in action

Four scanned pages of vertical Japanese from the 1917 first edition of Rashōmon (old-style kanji, historical kana, furigana) become horizontal Simplified Chinese through File Translation at high precision in reflow mode: the right-to-left reading order is kept, furigana stays out of the text, period terms such as Suzaku Avenue, ichime-gasa and momi-eboshi are kept, and all the body text on the 4 pages is translated.

A scanned old book page: the vertical title 羅生門 on the right and ten vertical columns of Japanese text to its left, with small furigana beside the kanji and page number 1 at the foot. — Page 1 of the 1917 first edition of Akutagawa's Rashōmon: a scan of vertical text in old-style kanji with furigana beside the characters. On the right, the opening of the translation in horizontal Simplified Chinese: the title and the first two paragraphs, the second of which runs on into page 2 of the book.

The opening of the translation: the bold title 罗生门 (羅生門) and two paragraphs of horizontal Simplified Chinese, beginning with the sentence about a servant waiting for the rain to stop under the Rashōmon gate. — Page 1 of the 1917 first edition of Akutagawa's Rashōmon: a scan of vertical text in old-style kanji with furigana beside the characters. On the right, the opening of the translation in horizontal Simplified Chinese: the title and the first two paragraphs, the second of which runs on into page 2 of the book.

Actual production output; 3 spots hand-corrected before display · Source: Rashōmon, Ryūnosuke Akutagawa, 1917 (The author, Ryūnosuke Akutagawa, died in 1927, more than 70 years ago; under Japanese copyright law, works of authors who died before the end of 1967 were protected for 50 years after death, so protection of this book expired at the end of 1977. The National Diet Library Digital Collections marks the item PDM (Public Domain Mark) and publishes it online (DOI 10.11501/1088339).)

Frequently asked questions

Can vertical Japanese books be processed?

Evaluate samples with the existing translation and extraction tools, checking reading order, missing characters and names against the source.

Can institutional records be uploaded directly?

First confirm processing rights and sensitivity, and read the current privacy and security information. This scenario page is not a certification of compliance.

Are institutional plans or private deployments included?

This page explains existing tools. It makes no commitment to procurement, institutional prices or private deployments; use the contact page for enquiries.

中文版

Geyi is a proprietary AI document platform owned and operated by Nayuta Technology and Culture Company Limited.