bikenshrestha / nlp-japanese-dictionary

More than just a simple dictionary software, NLP Japanese Dictionary integrates advanced natural language processing algorithms to help you read Japanese as effortlessly as possible.

NLP Japanese Dictionary

A Japanese - English dictionary software with OCR for Windows 7, 8, 10. (require .Net Framework 4.6.1)

More than just a simple dictionary, NLP Japanese Dictionary integrates advanced natural language processing algorithms to help you read Japanese as effortlessly as possible.

Main Features:

• Breakdown complicated sentences into simple words then give their reading, pronunciation, and definitions.
• Sort definitions of a word based on its context in a sentence.
• Include 170000+ Japanese - English entries from JMdict.
• Allow searching by Kana, Kanji, Romaji, and English.
• Provide Kanji information and their writing.
• Provide sentence examples.
• Extract text from images using OCR.
• Support text-to-speech for Japanese.
• Allow switching between day and night mode instantly.

Just extract and run NLPJapaneseDictionary.exe (require .Net Framework 4.6.1):

Website

3rd Party Open Source Projects

This application uses various open source projects (binary or source version). Please see CREDITS.md for details. The most important ones are:

• Japanese sentence segmentation: Kuromoji under Apache License 2.0 and Mecab-Ipadic.
• Japanese - English dictionary: Japanese-Multilingual Dictionary under Creative Commons Attribution-ShareAlike Licence 3.0.
• Kanji databases: KANJIDIC2, KanjiVG under Creative Commons Attribution-ShareAlike Licence 3.0.
• Example sentences: Tanaka Corpus and Tatoeba Project under Creative Commons Attribution Licence 2.0.

Project Structure

1. Jocr & Mqdf (.Net Standard 1.4)

Required nuget packets System.Numerics.Vectors and System.Threading.Tasks.Parallel.

These projects are used for Japanese OCR. They are written from scratch by us and do not rely on any other source code. Thus, you can easily build or use them in other projects. Please note that they are only trained for printing text.

Please see TestOCR project for general uses. To convert your images to Jocr.GrayImage format before processing, see NLPJapaneseDictionary.OCR.JorcImageConvert.BitmapToGrayImageJocr.

2. Kuromoji (.Net Standard 2.0)

This is a port from java to C# by us. For more details, please view the original project https://github.com/atilika/kuromoji.

3. NLPJapaneseDictionary WPF

Required nuget packets sqlite-net-pcl, Newtonsoft.Json, and System.Text.Encoding.CodePages.

The main project provides the user interface written in WPF (and WinForms).

Wiki

For normal users, please read our FAQ.

Contributing

We welcome any contributions from submitting bug reports to new feature recommendations.

For pull requests, please license your contributions under Apache License V2.0. We will retain your copyright and include your name in the CONTRIBUTORS.md file.

We plan to add a few white papers on how the OCR engine works so that you can help us improve it or write a new engine for other languages. Though we don't expect to finish these papers anytime soon. At the moment, the engine does not work well with uneven lighting images, text with noisy background, or skew text. Thus, preprocessing steps are really needed to be improved.

License

NLP Japanese Dictionary is licensed under the Apache License, Version 2.0. Please see LICENSE for details.