Batched character image processing
First Claim
1. For use in a process for recognizing characters on at least one document of a plurality of documents, the method comprising:
- scanning at least a portion of said document to produce scan data signals reflecting the presence of character elements at particular positions on the document;
storing said scan data signals;
applying a recognition algorithm to the stored scan data signals;
developing first identity signals representing each character recognized by said algorithm;
presenting simultaneously in side-by-side adjacent positions imaged of a group of characters which failed recognition by said algorithm to a specified confidence level and wherein a plurality of images of said group of characters are taken from different lines of the same document or from different documents;
said plurality of images of said group of characters being developed by scan data signals; and
determining by inspection the identity of characters failing recognition to the specified confidence level and represented by said simultaneously-presented images.
2 Assignments
0 Petitions
Accused Products
Abstract
Character recognition processing wherein each of a batch of documents is scanned to produce corresponding scan data signals forming a rectilinear data array of binary bits at the intersections of a rectangular coordinate grid. These signals are stored and processed by a recognition algorithm to produce identity signals for recognized characters. Groups of non-recognized characters are presented simultaneously to permit rapid identification by inspection. The identification of recognized characters is verified at high speed by simultaneously presenting the character images as respective groups sorted to have the same recognized identities.
38 Citations
33 Claims
-
1. For use in a process for recognizing characters on at least one document of a plurality of documents, the method comprising:
-
scanning at least a portion of said document to produce scan data signals reflecting the presence of character elements at particular positions on the document; storing said scan data signals; applying a recognition algorithm to the stored scan data signals; developing first identity signals representing each character recognized by said algorithm; presenting simultaneously in side-by-side adjacent positions imaged of a group of characters which failed recognition by said algorithm to a specified confidence level and wherein a plurality of images of said group of characters are taken from different lines of the same document or from different documents; said plurality of images of said group of characters being developed by scan data signals; and determining by inspection the identity of characters failing recognition to the specified confidence level and represented by said simultaneously-presented images. - View Dependent Claims (2, 3, 4, 5, 6, 7, 27, 29)
-
-
8. In a process for recognizing characters on at least one document of a plurality of documents, wherein at least a portion of said document is scanned to produce scan data signals reflecting the presence of character elements at particular locations on the document;
- said scan data signals being stored and operated on by a recognition algorithm to develop first identity signals representing each character recognized by the algorithm;
that improvement comprising the following steps; (1) presenting side-by-side adjacent images of a group of said characters at least some which are taken from different lines on the same document or from different documents, and which group of characters failed recognition to a specified confidence level by said algorithm, said character images being presented simultaneously and developed by scan data signals; (2) determining by inspection the identity of at least some of said characters failing recognition to the specified confidence level and represented by said simultaneously-presented images; and (3) developing second identity signals for the initially non-recognized but now-identified characters, said second identity signals serving with said first identity signals to develop at least part of an output text for said documents. - View Dependent Claims (9, 10, 11, 12, 28)
- said scan data signals being stored and operated on by a recognition algorithm to develop first identity signals representing each character recognized by the algorithm;
-
13. For use in a process for recognizing characters on at least one document wherein at least a portion of said document is scanned to produce and store scan data signals reflecting the presence of character elements at particular positions on the document;
- the characters represented by said stored scan data signals being initially analyzed by recognition procedures including that of processing the stored character scan data signals by a recognition algorithm;
the method of verifying the identity of characters recognized by said procedures comprising the following steps; (1) sorting the recognized characters into groups with a common characterization and without regard to the document where the character originated; (2) simultaneously presenting images of a number of characters of each of said groups respectively, said images being formed by the corresponding stored scan data signals; (3) determining by inspection the presence of any character in the displayed group failing to have the characterization common to that group and therefore incorrectly recognized; (4) determining by inspection the correct identity of such incorrectly-recognized character; (5) developing a corrected identity signal for each initially incorrectly-recognized but subsequently-identified character; and (6) utilizing said corrected identity signal to create at least part of an output text. - View Dependent Claims (14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 33)
- the characters represented by said stored scan data signals being initially analyzed by recognition procedures including that of processing the stored character scan data signals by a recognition algorithm;
-
26. In a process for recognizing a large number of characters from at least one document of a plurality of documents, wherein at least a portion of the document is scanned to produce scan data signals for characters to be recognized and reflecting the presence of character elements at particular positions on the document;
- said scan data signals being stored and operated on by a recognition algorithm to develop first identity signals representing each character recognized by the algorithm;
the method comprising the following steps; (1) sorting character images developed by scan data signals representing character images failing recognition by the algorithm to a specified confidence level; (2) simultaneously presenting side-by-side adjacent images of a number of said stored character images taken from different lines on the same document of from different documents and which had previously failed recognition to the specified confidence level; (3) determining by inspection the identity of at least some of said simultaneously-presented character images; and (4) developing second identity signals for the characters previously failing recognition to the specified confidence level but now-identified, said second identity signals being used with said first identity signals to develop at least part of an output text for said scanned characters.
- said scan data signals being stored and operated on by a recognition algorithm to develop first identity signals representing each character recognized by the algorithm;
-
30. In a process for recognizing characters on at least one of a plurality of documents wherein at least a portion of each document is scanned to produce scan data signals for characters to be recognized and reflecting the presence of character elements at particular locations on the document;
- said scan data signals being stored and operated on by a recognition algorithm;
the method comprising the following steps; (1) simultaneously presenting side-by-side adjacent images developed by scan data signals for characters originating from different lines on the same document or from different documents and which had previously failed to be recognized to a specified confidence level by said recognition algorithm; (2) determining by inspection the identity of at least some of said simultaneously-presented character images; and (3) developing an output text for said scanned characters including those recognized by said algorithm and those determined by said inspection.
- said scan data signals being stored and operated on by a recognition algorithm;
-
31. In a process for recognizing characters on a batch of documents wherein at least a portion of each document is scanned to produce scan data signals for characters to be recognized and reflecting the presence of character elements at particular locations on the document;
- said scan data signals being stored, and the stored scan data signal being analyzed by recognition procedures;
the method of verifying the identity of characters recognized by said procedures comprising the following steps; (1) sorting the characters recognized by said procedures into groups with a common characterization and without regard the document Where the character originated; (2) simultaneously presenting images of a number of the characters of each group with said images being formed by corresponding scan data signals; (3) determining by inspection of a presented group any characters which were incorrectly recognized; (4) determining by inspection on the correct identity of such incorrectly-recognized character; (5) developing by a corrected identity signal for each initially incorrectly-recognized but subsequently-identified character; and (6) utilizing said corrected identity signal to create at least part of an output text for said batch of documents.
- said scan data signals being stored, and the stored scan data signal being analyzed by recognition procedures;
-
32. In a process for recognizing characters in a batched plurality of documents wherein at least a portion of each document is scanned to produce scan data signals for characters to be recognized and reflecting the presence of character elements at particular locations on the document;
- said scan data signals being stored and operated on by a recognition algorithm;
the method comprising the following steps; (1) simultaneously presenting side-by-side adjacent images of a plurality of character images developed by scan data signals for characters originating from at least two of said documents an which previously failed to be recognized at a specified level of confidence by said recognition algorithm; (2) determining by inspection the identity of at least some of said simultaneously-presented character images; and (3) developing an output text for said scanned characters including those recognized by said algorithm and those determined by said inspection.
- said scan data signals being stored and operated on by a recognition algorithm;
Specification