Credits and Licences
Scan Chinese is built on the work of others. Several of the components below require attribution as a condition of their licence; this page is how we meet that obligation, and how we say thank you.
Dictionary data
- CC-CEDICT
- The Chinese–English dictionary behind every definition shown in the app. Copyright the CC-CEDICT contributors, licensed under Creative Commons Attribution-ShareAlike 4.0. cc-cedict.org
- Make Me a Hanzi
- Stroke data for handwriting recognition, by Shaunak Kishore, derived from the Arphic fonts and distributed under the Arphic Public License and LGPL. github.com/skishore/makemeahanzi
Recognition
- HanziLookup GPL v3
- Stroke-based handwriting recognition in the browser, by Gábor Ugray, based on the original HanziLookup by Jordan Kiang. Licensed under the GNU General Public License v3. Source: github.com/gugray/HanziLookupJS. The copy we serve is unmodified; the corresponding source is available at that address.
- PaddleOCR (PP-OCRv4)
- The neural recognition model that runs in your browser, by the PaddlePaddle authors, licensed under Apache 2.0. ONNX conversion via RapidOCR.
- Tesseract.js
- Browser OCR, licensed under Apache 2.0. tesseract.projectnaptha.com
- ONNX Runtime Web
- Runs the neural model in the browser. Copyright Microsoft, licensed under the MIT licence.
Language processing
- pypinyin
- Mandarin romanisation. MIT licence.
- PyCantonese
- Cantonese Jyutping and Yale romanisation, by Jackson Lee. MIT licence. pycantonese.org
- deep-translator
- Free translation fallback. MIT licence.
Typography and interface
- Noto Sans CJK
- Google's pan-CJK typeface, used to render the character templates the image matcher compares against. Licensed under the SIL Open Font License 1.1.
- JetBrains Mono
- The monospaced typeface in the interface, by JetBrains, licensed under the SIL Open Font License 1.1.
Location data
- DB-IP IP-to-Country Lite
- Used to work out the approximate country an account signed up from, entirely on our own server; no address is ever sent to a third party to be resolved. By DB-IP, licensed under Creative Commons Attribution 4.0. db-ip.com
Infrastructure
Flask, MariaDB, nginx, Postfix, OpenDKIM, PostSRSd, Pillow, NumPy and PyMySQL, each under their own open-source licences. Recognition, translation and speech synthesis are provided by Google Cloud under their terms.
Trademarks
Product names mentioned here belong to their respective owners and are used only to identify the software concerned.
If you believe something is credited incorrectly or is missing, please write to hello@scanchinese.com and we will correct it.