Character matching process for text converted from images
US6668085B1 · kind B1 · utility
Assignee
Inventor
Key dates
| Filing date | Aug 1, 2000 |
| Grant date | Dec 23, 2003 |
| Priority date | — |
| Expiry date | Feb 6, 2022 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06V30/1423
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
An improved method of deriving the correct text from text with errors converted from a character recognition device includes the need for significantly less human intervention for correction of the converted text. The method includes receiving as input a converted text sequence from a character recognition device, comparing a character sequence made up of one or more in-sequence characters of the converted text sequence to a first table containing either unidirectional or bi-directional substitution sequences to obtain a set of substitution sequences associated with the character sequence, and subsequently comparing the character sequence to a second table containing either unidirectional or bi-directional substitution sequences, wherein if the first table is a unidirectional table then the second table is a bi-directional table and if the first table is a bi-directional table then the second table is a unidirectional table, to obtain any additional possible substitution sequences associated with the character sequence, where the obtained character sequence and associated substitution sequences represent the set of possible text sequences for the character sequence of the converted…
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.