US5113453A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Character recognition method and apparatus
Assignee: France Telecom R&D SA
•
Filing: 1990-05-14
•
Granted: 1992-05-12
•
Expires: 2010-05-14
Abstract
Abstract An image system (SPI) provides an image file which is recorded in a memory (3). A segmentation and framing module (M1) isolates each of the characters in the image file within individual matrices. A module (M2) finds primitive shapes in each character matrix by initially finding strokes therein and then substracting the strokes from the matrix. A module (M3) then makes a description of primitive shapes in the form of at least one partially ordered list of primitive shapes, and a module (M4) compares said description with standard description expressed in the same manner in a data base (M5). Character recognition, per se, is performed by determining the distances between the description of a character to be identified and the standard descriptions.
Inventors
- 👤 Jean-Claude Simon
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-29 21:09
CPC Classifications
G; G06; G06V; G06V10/00; G06V10/40; G06V10/44; G06V30/00; G06V30/10
Claims Analysis
Total Claims: 18
Claim 1
Independent Claim
170 words
A method of processing digital signals representative of image lines, and in particular of characters, the method comprising the steps of: a) isolating each elementary shape and character, and framing said each shape and character in a matrix having a predetermined configuration, in particular a rectangular configuration; and b) performing a search inside the matrix for primitive shapes of predetermined types, said primitive shapes being referred to as first level primitive shapes; wherein operation b) comprises: b11) searching inside the matrix for quanta, with each quantum being defined as a homogeneous block having a minimum number of pixels having intensity greater than a local threshold; b12) detecting a plurality of sequences of overlapping quanta, with each sequence being representative of a stroke; and wherein adjacent strokes recognized in this way are subtracted from the matrix, thereby leaving characteristic zones containing said primitive shapes, wherein the first level primitive shapes comprises: the ends of strokes; changes in direction; cusps; crosses or branching points having at least three branches; and closed loops.
Claim 2
Independent Claim
170 words
A method of processing digital signals representative of image lines, and in particular of characters, the method comprising the steps of: a) isolating each elementary shape and character, and framing said each shape and character in a matrix having a predetermined configuration, in particular a rectangular configuration; and b) performing a search inside the matrix for primitive shapes of predetermined types, said primitive shapes being referred to as first level primitive shapes; wherein operation b) comprises: b11) searching inside the matrix for quanta, with each quantum being defined as a homogeneous block having a minimum number of pixels having intensity greater than a local threshold; b12) detecting a plurality of sequences of overlapping quanta, with each sequence being representative of a stroke; and wherein adjacent strokes recognized in this way are subtracted from the matrix, thereby leaving characteristic zones containing said primitive shapes, wherein the first level primitive shapes comprises: the ends of strokes; changes in direction; cusps; crosses or branching points having at least three branches; and closed loops.
Claim 3
Independent Claim
143 words
A method of processing digital signals representative of image lines, and in particular of characters, the method comprising the steps of: a) isolating each elementary shape and character, and framing said each shape and character in a matrix having a predetermined configuration, in particular a rectangular configuration; and b) performing a search inside the matrix for primitive shapes of predetermined types, said primitive shapes being referred to as first level primitive shapes; wherein operation b) comprises: b11) searching inside the matrix for quanta, with each quantum being defined as a homogeneous block having a minimum number of pixels having intensity greater than a local threshold; b12) detecting a plurality of sequences of overlapping quanta, with each sequence being representative of a stroke; and wherein adjacent strokes recognized in this way are subtracted form the matrix, thereby leaving characteristic zones containing said primitive shapes.
Claim 4
Independent Claim
143 words
A method of processing digital signals representative of image lines, and in particular of characters, the method comprising the steps of: a) isolating each elementary shape and character, and framing said each shape and character in a matrix having a predetermined configuration, in particular a rectangular configuration; and b) performing a search inside the matrix for primitive shapes of predetermined types, said primitive shapes being referred to as first level primitive shapes; wherein operation b) comprises: b11) searching inside the matrix for quanta, with each quantum being defined as a homogeneous block having a minimum number of pixels having intensity greater than a local threshold; b12) detecting a plurality of sequences of overlapping quanta, with each sequence being representative of a stroke; and wherein adjacent strokes recognized in this way are subtracted form the matrix, thereby leaving characteristic zones containing said primitive shapes.
Claim 5
Dependent Claim
109 words
A method according to claim 15, wherein step b11) comprises: scanning through the matrix by means of at least one predetermined scanning line which is moved parallel to itself in translation, with a current run being defined by a stack of pixels encountered by the scanning line having an intensity greater than a background intensity, said current run being considered as a first run at the beginning of scanning; searching for contacts being the current run, one or more upsteam runs, and also one or more downstream runs in the scanning direction; and recognizing that a quantum is present when a predetermined contact condition is satisfied between adjacent runs.
Claim 6
Dependent Claim
63 words
A method according to claim 1, further including: e) scanning the characteristic zones by means of at least one line having predetermined motion, in order to establish an ordered list of characteristic zones each having one of the primitive shapes identified therein; and f) comparing said list of primitive shapes obtained in this way with a set of character descriptions contained in memory.
Claim 7
Dependent Claim
30 words
A method according to claim 4, wherein the lists of primitive shapes are partially ordered, with primitive shapes which are encountered simultaneously during scanning being collected together in any order.
Claim 8
Dependent Claim
52 words
A method according to claim 5, wherein each matrix is subjected to at least two predetermined scans, which are associated with respective partially ordered lists, said lists being compared with partially ordered lists contained in memory, with each character or shape to be recognized being associated with at least one such lists.
Claim 9
Dependent Claim
21 words
A method according to claim 4, wherein the lists constituting invariant descriptions in the memory are classified by non-disjoint classes Ai.
Claim 10
Dependent Claim
33 words
A method according to claim 7, wherein the lists obtained are compared with the lists in a memory by calculating the distance between lists of characters to be compared within a class Ai.
Claim 11
Dependent Claim
42 words
A method according to claim 8, wherein, when the distance exceeds a threshold, the following steps are additionally performed: g) the character to be identified is eroded in order to reduce its thickness; and h) steps b) to f) are performed again.
Claim 12
Independent Claim
149 words
An apparatus for processing digital signals representative of image lines, and in particular of characters, the apparatus comprising: a) means for isolating each elementary shape and character, and means for framing said each shape and character in a matrix having a predetermined configuration, in particular a rectangular configuration; and b) means for searching inside the matrix for primitive shapes of predetermined types, said primitive shapes being referred to as first level primitive shapes; wherein said searching means comprises: b11) means for searching inside the matrix for quanta, with each quantum being defined as a homogeneous block having a minimum number of pixels having intensity greater than a local threshold; b12) means for detecting a plurality of sequences of overlapping quanta, with each sequence being representative of a stroke; b13) means for subtracting from the matrix adjacent strokes recognized in this way, thereby leaving characteristic zones containing said primitive shapes.
Claim 13
Independent Claim
149 words
An apparatus for processing digital signals representative of image lines, and in particular of characters, the apparatus comprising: a) means for isolating each elementary shape and character, and means for framing said each shape and character in a matrix having a predetermined configuration, in particular a rectangular configuration; and b) means for searching inside the matrix for primitive shapes of predetermined types, said primitive shapes being referred to as first level primitive shapes; wherein said searching means comprises: b11) means for searching inside the matrix for quanta, with each quantum being defined as a homogeneous block having a minimum number of pixels having intensity greater than a local threshold; b12) means for detecting a plurality of sequences of overlapping quanta, with each sequence being representative of a stroke; b13) means for subtracting from the matrix adjacent strokes recognized in this way, thereby leaving characteristic zones containing said primitive shapes.
Claim 14
Dependent Claim
108 words
Apparatus for processing digital signals representing image lines, to identify shapes and characters, as recited in claim 10, comprising: means for segmenting and framing characters, said means being suitable for receiving image data from an imaging system; means for recognizing primitive shapes by subtracting strokes; means for describing the primitive shapes as at least one list obtained which is in general partially ordered; means for storing character descriptions in the forms of predetermined reference lists; means for comparing the lists obtained with the reference lists; and means for identifying a character as matching a character which gives the smallest distance between the lists obtained and the reference lists.
Claim 15
Dependent Claim
27 words
Apparatus according to claim 11 further comprising: fining means for reducing the thickness of a character prior to reapplying it to the means for recognizing primitive shapes.
Claim 16
Independent Claim
270 words
An apparatus for processing digital signals representative of image lines, and in particular of characters, the apparatus comprising: a) means for isolating each elementary shape and character, and means for framing said each shape and character in a matrix having a predetermined configuration, in particular a rectangular configuration; and b) means for searching inside the matrix for primitive shapes of predetermined types, said primitive shapes being referred to as first level primitive shapes; wherein said searching means comprises: b11) means for searching inside the matrix for quanta, with each quantum being defined as a homogeneous block having a minimum number of pixels having intensity greater than a local threshold; b12) means for detecting a plurality of sequences of overlapping quanta, with each sequence being representative of a stroke; b13) means for subtracting form the matrix adjacent strokes recognized in this way, thereby leaving characteristic zones containing said primitive shapes; (c) means for segmenting and framing characters, said means being suitable for receiving image data from an imaging system; (d) means for recognizing primitive shapes by subtracting strokes; (e) means for describing the primitive shapes as at least one list obtained which is in general partially ordered; (f) means for storing character descriptions in the forms of predetermined parenthesized reference lists; (g) means for comparing the lists obtained with the reference lists; and (h) means for identifying a character as matching the character which gives the smallest distance between the lists obtained and the reference lists, wherein the primitive shapes comprise: the ends of strokes; changes in direction; cusps; crosses or branching points having at least three branches; and closed loops.
Claim 17
Independent Claim
270 words
An apparatus for processing digital signals representative of image lines, and in particular of characters, the apparatus comprising: a) means for isolating each elementary shape and character, and means for framing said each shape and character in a matrix having a predetermined configuration, in particular a rectangular configuration; and b) means for searching inside the matrix for primitive shapes of predetermined types, said primitive shapes being referred to as first level primitive shapes; wherein said searching means comprises: b11) means for searching inside the matrix for quanta, with each quantum being defined as a homogeneous block having a minimum number of pixels having intensity greater than a local threshold; b12) means for detecting a plurality of sequences of overlapping quanta, with each sequence being representative of a stroke; b13) means for subtracting form the matrix adjacent strokes recognized in this way, thereby leaving characteristic zones containing said primitive shapes; (c) means for segmenting and framing characters, said means being suitable for receiving image data from an imaging system; (d) means for recognizing primitive shapes by subtracting strokes; (e) means for describing the primitive shapes as at least one list obtained which is in general partially ordered; (f) means for storing character descriptions in the forms of predetermined parenthesized reference lists; (g) means for comparing the lists obtained with the reference lists; and (h) means for identifying a character as matching the character which gives the smallest distance between the lists obtained and the reference lists, wherein the primitive shapes comprise: the ends of strokes; changes in direction; cusps; crosses or branching points having at least three branches; and closed loops.
Claim 18
Dependent Claim
28 words
Apparatus according to claim 13, further including: fining means suitable for reducing the thickness of a character prior to reapplying it to the means for recognizing primitive shapes.
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) | 1995-05-12 | 1995-11-11 | 1996-05-12 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1999-05-12 | 1999-11-11 | 2000-05-12 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2003-05-12 | 2003-11-11 | 2004-05-12 | $7700.00 | Window Closed / Lapsed |