Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
CHARACTER RECOGNITION UNIT
Document Type and Number:
Japanese Patent JPH04318687
Kind Code:
A
Abstract:

PURPOSE: To discriminate the type of a character and to output a correct character code without being affected by a similar character within a single character type or between multicharacter types under the condition that a correct solution is obtained among candidates from dictionaries for each character type.

CONSTITUTION: In a processing section 23, a keyword is detected from the character code string obtained by recognizing a character string consisting of a plurality of characters as English alphabetic characters, numeric characters, Kana characters, and symbols, similar characters and non-similar characters are discriminated in the character string, a character code for non-similar character is identified, a character code for a similar character that could not be identified is identified, and based on the result a character code string is re- constructed. From this, under the condition that a correct solution is obtained among candidates selected from dictionaries for each character type, a character type is discriminated without being affected by similar character within a single character type and between multicharacter types, and a correct character code can be output.


Inventors:
FUTAGAWA KATSUMI
Application Number:
JP8536591A
Publication Date:
November 10, 1992
Filing Date:
April 17, 1991
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
NTT DATA TSUSHIN KK
International Classes:
G06K9/46; G06K9/62; G06K9/68; G06F17/30; G06K9/72; (IPC1-7): G06F15/40; G06K9/62
Attorney, Agent or Firm:
Akita Haruki



 
Previous Patent: JPS4318686

Next Patent: MEMORY CARD EXTRACTING MECHANISM