# US 5,703,972 A · Certifiable optical character recognition > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 5,703,972 A - Document ID: US5703972A - Status: Granted - Kind: A - Application number: 08/560,299 - Filed: 1995-11-17 - Granted: 1997-12-30 - Earliest claimed: 1992-10-09 - Assignees: Panasonic Technologies, Inc. - Inventors: Daniel P. Lopresti; Henry F. Korth; Jonathan S. Sandberg; Richard J. Lipton - Web page: https://patentgrep.com/patent/US5703972A ## CPC - G06V 30/12: Detection or correction of errors, e.g. by rescanning the pattern - G06V 30/1444: {Selective acquisition, locating or processing of specific regions, e.g. highlighted text, fiducial marks or predetermined fields} - G06V 30/10: Character recognition - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract A certificate, calculated on the text of a document to be reproduced, is included in the document in a form which is machine-readable. Upon optical character recognition scanning, or other digitizing reproduction, the certificate is also scanned and stored for comparison. A new certificate is calculated on the contents of the reproduced document and compared to the scanned certificate. If the two certificates are different, it is known that at least one error exists in the reproduced document. More than one certificate can be provided for a document in order to facilitate error location on the document and error correction therein. ## Claims ### Claim 1 An information medium adapted to be scanned by a machine to recover a machine representation of a spatial relationship of blocks in said medium, comprising: a plurality of machine-rendered blocks arranged in said medium including at least one textblock of multiple lines of human recognizable characters and another block containing human recognizable information; and at least one non-textual machine-readable marker representative, at least in part, of a machine-representation of a spatial relationship of said machine-rendered blocks in said medium machine-derived when said blocks are machine-rendered and encoding said spatial relationship of said blocks in said medium to permit said machine representation of said spatial relationship to be recovered. ### Claim 2 (depends on 1) The information medium of claim 1 wherein said marker is machine-derived from information content of one or more of said blocks. ### Claim 3 (depends on 2) The information medium of claim 2 wherein said marker encodes information identifying at least one of said blocks as containing text, figure, photograph or equation content. ### Claim 4 (depends on 1) The information medium of claim 1 wherein said marker is further machine-derived from original machine-represented character data representing said human recognizable characters in said at least one textblock to permit errors in said at least one scanned textblock of human recognizable characters to be identified based on a comparison between said marker and at least one marker derived from said at least one scanned textblock of human recognizable characters. ### Claim 5 (depends on 1) The information medium of claim 1 wherein said marker is machine-derived from original machine-represented character data representing said human recognizable characters in said at least one textblock to permit error-corrected machine-represented character data to be recovered by identifying errors in said scanned human recognizable characters based on a comparison between said marker and at least one marker derived from said scanned human recognizable characters. ### Claim 6 (depends on 1) The information medium of claim 1 wherein said marker encodes information including the coordinates of at least one of said blocks on said information medium. ### Claim 7 (depends on 1) The information medium of claim 1 wherein said blocks include at least one block having image content. ### Claim 8 An information medium adapted to be scanned by a machine which stores error corrected machine-represented character data from original human recognizable characters in said medium, comprising: a plurality of lines of original human recognizable characters machine-rendered from original machine-represented character data; and at least one non-textual machine-readable marker machine-derived when said lines are machine rendered and encoded for use by a machine for identifying errors in said scanned original human recognizable characters to permit said original machine-represented character data to be recovered. ### Claim 9 (depends on 8) The information medium of claim 8 wherein said marker is derived by computing an error correcting code (ECC) based on each character in a text line of said original human recognizable characters. ### Claim 10 An information medium adapted to be scanned by a machine to recover a machine representation of a spatial relationship of blocks in said medium, comprising: a plurality of machine-rendered blocks arranged in said medium including at least one block having image content; and at least one non-textual machine-readable marker machine-derived when said blocks are machine-rendered and encoding said spatial relationship to permit said machine representation of said spatial relationship to be recovered. ### Claim 11 A system for permitting a machine representation of a spatial relationship of blocks in a recorded information medium to be recovered by scanning said medium, comprising: a recorded information medium adapted to be scanned by a machine, said medium including a plurality of machine-rendered blocks arranged in said medium in accordance with said machine representation of said spatial relationship, said blocks including at least one textblock of human recognizable characters, and at least one non-textual machine-readable marker, said marker being machine-derived when said blocks are machine rendered and encoding said spatial relationship of blocks in said medium; means for scanning said information medium to generate scanned block data and scanned marker data; and means for applying said scanned marker data to said scanned block data to recover said machine representation of said spatial relationship. ### Claim 12 A system for permitting error-corrected machine-represented character data to be recovered from a recorded information medium by scanning said medium, comprising: a recorded information medium adapted to be scanned by a machine, said medium including a plurality of lines of human recognizable characters, said lines being machine-rendered from original machine-represented character data, said medium further including at least one non-textual machine-readable marker, said marker being machine-derived when said lines are machine-rendered and encoded for use by a machine in identifying errors in said scanned human recognizable characters; means for scanning said information medium to generate scanned character data and scanned marker data; and means for applying said scanned marker data to said scanned character data to identify errors in said scanned character data, thereby permitting said error-corrected machine-represented character data to be recovered. ### Claim 13 A system for permitting error-corrected machine-represented character data to be recovered from a recorded information medium adapted to be scanned by a machine, said medium containing a plurality of machine-rendered lines of human recognizable characters, and at least one non-textual machine-readable marker, said marker being machine-derived when said lines are machine-rendered and encoded for use by a machine in identifying errors in said scanned human recognizable characters, said system comprising: means for scanning said information medium to generate scanned character data and scanned marker data; and means for applying said scanned marker data to said scanned character data to identify errors in said scanned character data, thereby permitting said error-corrected machine-represented character data to be recovered.