WINTONE TH-OCR
“delivering the best Asian languages recognition technologies”
Products

Embedded Handwriting Recognition

Purchase now
Overview

WINTONE handwriting recognition technology is another award-winning professional technology specially designed for embedding handwriting input recognition into mobile devices. It also provides advanced handwriting recognition core to be up to over 95%, which not only processes complex Simplified and Traditional Chinese handwriting, but also recognizes English, Japanese, Korean, Numbers, and Punctuations and so on.

Main Features
 
  • ■　The speed of recognition process (testing results from WINTONE)
  •     ■　2 words/per second (PalmOS 3.0，EZ328(16MHz)
  •     ■　4 words/ps (PalmOS 4.0，DragonBall EZ328(33MHz)
  •     ■　4 words/ps (WinCE 3.0，SA1100(206MHz))
  • ■　Memory requirements
  •     ■　Stack: <1K Bytes
  •     ■　Dynamic Memory: <2K Bytes
  • ■　High-accuracy handwriting recognition
  •     ■　> 95% (be related to the quality of handwriting)
  • ■　To be able to process multiple languagesrecognition
  •     ■　English
  •     ■　Numbers
  •     ■　Simplified Chinese
  •     ■　Traditional Chinese
  •     ■　Japanese
  •     ■　Korean
  •     ■　Arabic
  •     ■　... ...
 
Supported CPU Types

  • ■　ARM
  • ■　X86
  • ■　MIPS
  • ■　Epson
  • ■　Philips TriMedia 
  • ■　Motorola Dragon Ball EZ328
  • ■　Other 16 or 32 bits CPU，We need compiler and will finish migrate in five days.
 
Recognition Range:

  • ■　GB18030 (27484 characters)
  • ■　GBK (21003 characters)
  • ■　GB2312 (6763 characters)
  • ■　BIG5 (13053 characters)
  • ■　Standard capital or lowercase, English letters, Arabic numerals

Size:

500K (GB I)-900K (GBK)

Output Code

  • ■　GB2312
  • ■　UNICODE
 
System requirements:

    ■　Linux
  • ■　Symbian
  • ■　Windows Mobile
  • ■　PalmOS
  • ■　SDS
  • ■　HOPEN
  • ■　VxWorks
  • ■　Nucleus
  • ■　PPSM
  • ■　……
 
Our successfully partners

WINTONE handwriting input system is an ideal solution so that can be
widely combined with PDA, intelligent mobile devices and the like. Also,
its powerful OCR recognition core has been demonstrated by its more
partners, such as Samsung, NEC, Bird, CECT, Panda, Epson, Kejian
and so on.