Method for identifying unrecognizable characters in optical character recognition machines

• US 4,914,709 A
• Filed: 06/02/1989
• Issued: 04/03/1990
• Est. Priority Date: 06/02/1989
• Status: Expired due to Term

• Current Assignee: Global 360, Inc. (Open Text Corporation)
• Original Assignee: Eastman Kodak Company
• Inventors: Rudak, Peter
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Jung, Yon
• Application Number:
  US07/360,967
• Time in Patent Office: 305 Days
• Field of Search: 382/57, 382/7, 382/65, 382/44, 358/22, 358/180, 358/183, 235/437
• US Class Current: 382/311
• CPC Class Codes: G06V 10/987 with the intervention of an operator

Abstract

A method for identifying a character which cannot be machine read so that the operator may observe and hopefully recognize the character in question. A bit-map video image of the unrecognized character(s) is inserted in the ASCII data line of neighboring characters to create an impression of the original line of text from the document. A data entry operator uses this information to enter the required correct character(s) via the keyboard or other means. This reject/reentry method allows for quick operator response, and minimizes data storage and transmission of video information.

Claims

1. A method of processing documents having characters encoded thereon in a document reentry system comprising the steps of:

moving the documents along a path in serial flow;

reading data from said documents during said movement;

interpreting the data read and converting said data into readable characters;

establishing a threshold level of certainty below which the character is deemed unrecognizable;

identifying readable and unrecognizable characters on the basis of said threshold level;

displaying said readable characters in lines as they appear on the document;

generating a bit mapped image of each unrecognizable character;

inserting said bit mapped image of the unrecognizable character in place of the unrecognizable character on said display within a line of correctly read characters being adjacent thereto;

entering the correct character after viewing said bit mapped image in context with said readable characters.

2. A method of processing documents having characters encoded thereon in a document reentry system comprising the steps of:

moving the documents along a path in serial flow;

reading data from said documents during said movement;

identifying certain fields on said document for processing using optical character recognition techniques;

storing in a memory the data read from said fields;

determining whether the encoded characters read from said fields are read correctly or deemed unrecognizable with respect to a predetermined criteria and generate signals indicating the position within a line an unrecognizable character;

displaying the line of readable characters from the document using a standardized font;

extracting from said memory the coordinates on the document of a bit mapped image of the unrecognizable character;

inserting said bit mapped image in place of said unrecognizable character in the line of correctly read characters; and

entering the correct character after viewing.

3. A method of processing documents having characters encoded thereon in a document reentry system comprising the steps of:

scanning each document line by line and pixel by pixel;

identifying certain fields on said documents for processing using optical character recognition techniques;

storing the scanned data from said fields;

determining when the encoded characters read from said fields are read correctly or deemed unrecognizable with respect to a predetermined criteria and generate signals indicating the position within a line having unrecognizable characters;

extracting from said memory the coordinates on said document of a bit mapped image of each of said unrecognizable characters;

displaying the line of correctly read characters when said line has been indicated as containing at least one unrecognizable character; and

inserting said bit mapped image in the position of each of said unrecognizable characters in said line of correctly read characters thereby displaying the unrecognizable in context with adjacent correctly read characters.

4. The method of processing documents as set forth in claim 3 in which said extraction step includes obtaining the Y-offset, X-offset on said document and the height and width of the smallest rectangle capable of surrounding each of said unrecognizable characters.

5. The method of processing documents as set forth in claim 4 in which said extraction steps further include a masking step that masks out any marks on either side of the bit mapped image such as portions of correctly read characters not associated with each of said unrecognizable characters.

6. The method of processing documents as set forth in claim 3 which further includes a storage step for storing the line of readable characters and the bit mapped image of each of said unrecognizable characters for display and correction at a later date.

7. The method of processing documents as set forth in claim 3 in which said display step displays a line of correctly read characters using a standard font.

8. The method of processing elements as set forth in claim 7 in which said display step displays a line of correctly read characters in ASCII format.

9. The method of processing documents as set forth in claim 5 in which said masking step effects only the left-most and right-most groups of pixels of the bit mapped image.

10. The method of processing documents as set forth in claim 5 in which a byte boundary is determined to be the smallest multiple of 8 pixels.

11. The method of processing documents as set forth in claim 10 in which the masking step further includes rounding the X-offset to the next smallest byte boundary and then subtracting the magnitude of said byte boundary from the X-offset and storing the result as a remainder.

12. The method of processing documents as set forth in claim 11 in which the masking step further includes adding said remainder to said width and rounding the result to the next higher byte boundary to obtain the byte-boundary width.

13. The method of processing documents as set forth in claim 12 in which the masking step further includes the use of a masking byte which contains a logic "0" at all locations requiring masking and a logic "1" in other locations in that byte.

14. The method of processing documents as set forth in claim 13 in which the masking step further includes performing a logic AND operation with the video data and the masking byte to form a masked video byte.