ianchen06 / jlect-jhr Goto Github PK

#JLect Japanese Handwriting Recognition Tool (JLect JHR) (c) Zachary Read | JLect.com 2013-2016

##What is it? JLect JHR is a JavaScript library that provides support for Japanese handwritten character recognition. Users can draw a character in a defined HTML canvas element and the library will output guesses based on stroke direction.

##License The library and all associated files provided in the official repository are covered by the following Creative Commons license:

• Attribution-ShareAlike 3.0 Unported (CC BY-SA 3.0): https://creativecommons.org/licenses/by-sa/3.0/

##Example For a live example of the library in use, you can visit https://www.jlect.com and click on the "字 (draw)" button beside the search bar. A simplified example is also provided in the file "example.html" found in the official repository https://github.com/ZacharyRead/jlect-jhr.

##Advantages and disadvanges Pros:

• The library is open-sourced, so anyone can contribute to the project to improve it and anyone can use it as long as the license conditions are met.
• The library has no dependencies, which makes it easy to deploy in comparison to other tools.
• The library can be easily modified and expanded. The "example.html" file even provides data output that you can use to expand the existing dataset.
• The library, while being stroke-oriented, considers the fact that users are likely to draw characters in the wrong stroke order.

Cons:

• The library is stroke-oriented and does not handle cursive character input.
• The dataset is incomplete, so it may fail at recognizing a large number of characters.
• There is currently no touch support, so it will not work on mobile, tablet or other touch-screen-based devices.
• It only supports drawing one character at a time.

##Contributing I strongly encourage anyone who uses the library to contribute back to the project: https://github.com/ZacharyRead/jlect-jhr.