Patent US4783835A successfully refreshed from Google Patents.
US4783835A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method of character recognition
Assignee: Sumitomo Electric Industries Ltd
•
Filing: 1987-12-28
•
Granted: 1988-11-08
•
Expires: 2007-12-28
Abstract
Abstract An improved apparatus and method of recognizing a character is disclosed. The side frame structures of characters or figures are discriminated by left and right margin detectors and corresponding left and right side frame recognition circuits. A final character discrimination circuit synthesized both the results of a conventional character recognition unit and the frame recognition circuits, and permits discrimination between similarly shaped characters.
Inventors
- 👤 Koji Satoh
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-01 11:16
CPC Classifications
G; G06; G06V; G06V30/00; G06V30/10; G06V30/18; G06V30/182; G06V30/18162
Claims Analysis
Total Claims: 20
Claim 1
Independent Claim
399 words
In an apparatus for recognizing a character with a hand held scanner having an image sensor composed of a two-dimensional array of photoelectric elements, projection lamps, a focusing lens, and a binary conversion circuit for converting output analog signals of the photoelectric elements into digital signals to form a binary pattern of the character scanned, wherein the binary pattern of the character scanned is defined in a rectangular character area of rows and columns of cells in which digital bits representing the character scanned occur or do not occur, and said apparatus including character area determining means for determining the character area by identifying the rows and columns of cells having digital bits representing the character scanned, and character recognition means for abstracting the presence of absence of digital bits on each line of the identified rows or columns of the character area into a set of line features, identifying the set of abstracted line features of the scanned character within the character area, and producing a character recognition result therefrom, the improvement comprising: character margin detection means, receiving the outputs from said binary conversion circuit and said character area determining means, for detecting the presence or absence of digital bits occurring in at least one line proximate each of a left side margin and a right aide margin of the character area, and providing left and right side margin output signals indicative of the detected digital bits in each respective line of the left and right side margins; character frame recognizing means receiving the output signals from said character margin detection means for recognizing a left side frame structure, from the output signals for the detected line of the left side margin, and a right side frame structure, from the output signals for the detected line of the right side margin, based upon an order of occurrence of said left and right side margin output signals of said character margin detection means; and character discrimination synthesis means receiving a character recognition output from said character recognition means and a left and right side frame structure output from said character frame recognizing means for synthesizing the results of said character recognition means and of said character frame recognizing means so as to discriminate between similarly shaped characters on the basis of differences detected in their left and right side margins and to provide a final character identification based thereon.
Claim 2
Dependent Claim
399 words
In an apparatus for recognizing a character with a hand held scanner having an image sensor composed of a two-dimensional array of photoelectric elements, projection lamps, a focusing lens, and a binary conversion circuit for converting output analog signals of the photoelectric elements into digital signals to form a binary pattern of the character scanned, wherein the binary pattern of the character scanned is defined in a rectangular character area of rows and columns of cells in which digital bits representing the character scanned occur or do not occur, and said apparatus including character area determining means for determining the character area by identifying the rows and columns of cells having digital bits representing the character scanned, and character recognition means for abstracting the presence of absence of digital bits on each line of the identified rows or columns of the character area into a set of line features, identifying the set of abstracted line features of the scanned character within the character area, and producing a character recognition result therefrom, the improvement comprising: character margin detection means, receiving the outputs from said binary conversion circuit and said character area determining means, for detecting the presence or absence of digital bits occurring in at least one line proximate each of a left side margin and a right aide margin of the character area, and providing left and right side margin output signals indicative of the detected digital bits in each respective line of the left and right side margins; character frame recognizing means receiving the output signals from said character margin detection means for recognizing a left side frame structure, from the output signals for the detected line of the left side margin, and a right side frame structure, from the output signals for the detected line of the right side margin, based upon an order of occurrence of said left and right side margin output signals of said character margin detection means; and character discrimination synthesis means receiving a character recognition output from said character recognition means and a left and right side frame structure output from said character frame recognizing means for synthesizing the results of said character recognition means and of said character frame recognizing means so as to discriminate between similarly shaped characters on the basis of differences detected in their left and right side margins and to provide a final character identification based thereon.
Claim 3
Dependent Claim
46 words
The character recognition apparatus of claim 1, wherein said character frame recognizing means includes left and right side frame recognizing means for recognizing left and right side frame features in response to the orders of occurrence of said left and right side margin output signals, respectively.
Claim 4
Dependent Claim
40 words
The character recognition apparatus of claim 2, wherein said character frame recognizing means further comprises character frame discrimination means for discriminating a single frame construction for the binary pattern in response to said recognized left and right side frame features.
Claim 5
Dependent Claim
106 words
The character recognition apparatus of claim 1, wherein said character recognition means comprises character area determination means for determining the character area of the binary pattern of the scanned character through vertical and horizontal segmentation of the binary pattern, character segment feature extraction means for extracting character segment features of rows or columns of the binary pattern and for providing a feature code for each row or column segment feature, partial character feature extracting means for compressing said character segment feature codes into partial character feature codes, and character determination means for determining an identity of the character according to the said partial character feature codes.
Claim 6
Dependent Claim
66 words
The character recognition apparatus of claim 1, wherein said character frame recognizing means comprises a read only memory and a register, the content of the read only memory being read out to the register in response to address inputs, and an output signal of the register and said margin output signals of said character margin detection means constituting said address inputs to said read only memory.
Claim 7
Dependent Claim
67 words
The character recognition apparatus of claim 1, wherein said character recognition means includes character area determination means for determining the character area for the binary pattern of the scanned character by providing segmentations of the rows and columns of the binary pattern, and further wherein said character frame recognizing means recognizes the character frame feature from said margin output signals within said row segmentation and column segmentation.
Claim 8
Dependent Claim
33 words
The character recognition method of claim 1, wherein said step of providing margin output signals includes providing left and right side margin output signals from left and right side margin detecting means, respectively.
Claim 9
Dependent Claim
61 words
The character recognition method of claim 7, wherein said step of recognizing a frame feature includes recognizing left and right side frame features of the binary pattern by left and right side frame recognizing means in response to the orders of occurrence of said left and right side margin output signals of said left and right side margin detecting means, respectively.
Claim 10
Dependent Claim
35 words
The character recognition method of claim 8, further comprising the step of recognizing a single frame construction for the binary pattern by combining the left and right side frame features by character frame discrimination means.
Claim 11
Dependent Claim
26 words
The character recognition apparatus of claim 1, wherein at least said image sensor, projection lamps, and focusing lens are incorporated in a hand held scanner unit.
Claim 12
Dependent Claim
383 words
In a method for recognizing a character by scanning with a hand held scanner having an image sensor composed of a two-dimensional array of photoelectric elements, projection lamps, a focusing lens, and a binary conversion circuit which converts output analog signals of the photoelectric elements into digital signals to form a binary pattern of the character scanned, wherein the binary pattern of the character scanned is defined in a rectangular character area of rows and columns of cells in which digital bits representing the character scanned occur or do not occur, and by determining the character area by character area determining means for identifying the rows and columns of cells having digital bits representing the character scanned, and by processing the binary pattern with character recognition means for identifying line features on each line of the scanned character within said character area and producing a character recognition result, the improvement comprising: receiving the outputs from said binary conversion circuit and said character area determining means, detecting by character margin detection means the presence or absence of digital bits occurring in at least one line proximate each of a left side margin and a right side margin of the character area, and providing left and right side margin output signals indicative of the detected digital bits in each respective line of the left and right side margins within said character area; receiving the output signals from said character margin detection means, and recognizing a left side frame structure, from the output signals for the detected line of the left side margin, and a right side frame structure, from the output signals for the detected line of the right side margin, by character frame recognizing means in response to an order of occurrence of said left and right side margin output signals of said character margin detection means; and receiving a character recognition output from said character recognition means and a left and right side frame structure output from said character frame recognizing means, and synthesizing by character discrimination synthesis means the results of said character recognition means and of said character frame recognizing means so as to discriminate between similarly shaped characters on the basis of differences detected in their left and right side margins and to provide a final character identification based thereon.
Claim 13
Dependent Claim
383 words
In a method for recognizing a character by scanning with a hand held scanner having an image sensor composed of a two-dimensional array of photoelectric elements, projection lamps, a focusing lens, and a binary conversion circuit which converts output analog signals of the photoelectric elements into digital signals to form a binary pattern of the character scanned, wherein the binary pattern of the character scanned is defined in a rectangular character area of rows and columns of cells in which digital bits representing the character scanned occur or do not occur, and by determining the character area by character area determining means for identifying the rows and columns of cells having digital bits representing the character scanned, and by processing the binary pattern with character recognition means for identifying line features on each line of the scanned character within said character area and producing a character recognition result, the improvement comprising: receiving the outputs from said binary conversion circuit and said character area determining means, detecting by character margin detection means the presence or absence of digital bits occurring in at least one line proximate each of a left side margin and a right side margin of the character area, and providing left and right side margin output signals indicative of the detected digital bits in each respective line of the left and right side margins within said character area; receiving the output signals from said character margin detection means, and recognizing a left side frame structure, from the output signals for the detected line of the left side margin, and a right side frame structure, from the output signals for the detected line of the right side margin, by character frame recognizing means in response to an order of occurrence of said left and right side margin output signals of said character margin detection means; and receiving a character recognition output from said character recognition means and a left and right side frame structure output from said character frame recognizing means, and synthesizing by character discrimination synthesis means the results of said character recognition means and of said character frame recognizing means so as to discriminate between similarly shaped characters on the basis of differences detected in their left and right side margins and to provide a final character identification based thereon.
Claim 14
Dependent Claim
107 words
The character recognition method of claim 11, wherein said step of processing the binary pattern with character recognition means includes the steps of: determining the character area of the binary pattern of the scanned character through vertical and horizontal segmentation of the binary pattern; extracting character segment features for the rows or columns of the binary pattern and providing corresponding feature codes for the row or column segment features; extracting partial character features by compressing said character segment feature codes into partial character feature codes; and determining an identity of the character by character determination means in accordance with the occurrence of said partial character feature codes.
Claim 15
Dependent Claim
84 words
The character recognition method of claim 11, wherein said step of recognizing a frame feature of the binary pattern is obtained by said character frame recognizing means being comprised of a read only memory and the register in which the content of the read only memory is read out to the register in accordance with address inputs, and an output signal of the register and said margin output signals of said character margin detection means constitute the address inputs to said read only memory.
Claim 16
Dependent Claim
28 words
The character recognition method of claim 11, wherein said margin output signals of said character margin detection means represent an adjacent plurality of segments of the binary pattern.
Claim 17
Dependent Claim
60 words
The character recognition method of claim 11, wherein said step of processing the binary pattern with character recognition means includes determining a character area of the binary pattern of the character scanned through row segmentation and column segmentation, and said steps of providing margin output signals and recogniznng a frame feature are obtained within said row segmentation and column segmentation.
Claim 18
Dependent Claim
29 words
The character recognition method of claim 11, wherein said frame feature includes frame features of left and right sides and of top and bottom sides of the binary pattern.
Claim 19
Dependent Claim
27 words
The character recognition method of claim 11, wherein said frame feature includes frame features of left and right sides and a top side of the binary pattern.
Claim 20
Dependent Claim
23 words
The character recognition method of claim 11, wherein said frame feature includes frame features of left and right sides of the binary pattern.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1991-11-08 | 1992-05-09 | 1992-11-08 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1995-11-08 | 1996-05-09 | 1996-11-08 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 1999-11-08 | 2000-05-09 | 2000-11-08 | $7700.00 | Window Closed / Lapsed |