US5491758A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Automatic handwriting recognition using both static and dynamic parameters
Assignee: International Business Machines Corp
•
Filing: 1993-01-27
•
Granted: 1996-02-13
•
Expires: 2013-01-27
Abstract
Abstract Methods and apparatus are disclosed for recognizing handwritten characters in response to an input signal from a handwriting transducer. A feature extraction and reduction procedure is disclosed that relies on static or shape information, wherein the temporal order in which points are captured by an electronic tablet may be disregarded. A method of the invention generates and processes the tablet data with three independent sets of feature vectors which encode the shape information of the input character information. These feature vectors include horizontal (x-axis) and vertical (y-axis) slices of a bit-mapped image of the input character data, and an additional feature vector to encode an absolute y-axis displacement from a baseline of the bit-mapped image. It is shown that the recognition errors that result from the spatial or static processing are quite different from those resulting from temporal or dynamic processing. Furthermore, it is shown that these differences complement one another. As a result, a combination of these two sources of feature vector information provides a substantial reduction in an overall recognition error rate. Methods to combine probability scores from dynamic and the static character models are also disclosed.
Inventors
- 👤 Jerome R. Bellegarda
- 👤 David Nahamoo
- 👤 Krishna S. Nathan
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 13:31
CPC Classifications
G; G06; G06V; G06V30/00; G06V30/10; G06V30/32; G06V30/36; G06V30/373; G06F; G06F18/00; G06F18/20; G06F18/25; G06F18/254; G06F18/256; G06V30/19; G06V30/191; G06V30/1918; G06V30/333; G06V30/347
Claims Analysis
Total Claims: 10
Claim 1
Independent Claim
222 words
Handwriting recognition apparatus, comprising: handwriting transducer means having an output providing x-y coordinate information generated by a writer; first means, having an input coupled to said output of said handwriting transducer means, for partitioning the x-y coordinate information into first partitions, and for extracting and outputting temporally-based feature vectors from the first partitions of the x-y coordinate information; second means, having an input coupled to said output of said handwriting transducer means, for partitioning the x-y coordinate information into second partitions, and for extracting and outputting spatially-based feature vectors from the second partitions of the x-y coordinate information; and processing means, having a first input coupled to an output of said first means and a second input coupled to an output of said second means, for processing the temporally-based feature vectors that are output from said first extracting means and the spatially-based feature vectors that are output from said second extracting means to determine first probabilities that the temporally-based feature vectors represent predetermined first prototype characters and to determine second probabilities that the spatially-based feature vectors represent predetermined second prototype characters, said processing means further including means for determining, based on a weighted combination of said first probabilities and said second probabilities, a third probability that the x-y coordinate information generated by the writer represents one of a plurality of predetermined characters.
Claim 2
Dependent Claim
222 words
Handwriting recognition apparatus, comprising: handwriting transducer means having an output providing x-y coordinate information generated by a writer; first means, having an input coupled to said output of said handwriting transducer means, for partitioning the x-y coordinate information into first partitions, and for extracting and outputting temporally-based feature vectors from the first partitions of the x-y coordinate information; second means, having an input coupled to said output of said handwriting transducer means, for partitioning the x-y coordinate information into second partitions, and for extracting and outputting spatially-based feature vectors from the second partitions of the x-y coordinate information; and processing means, having a first input coupled to an output of said first means and a second input coupled to an output of said second means, for processing the temporally-based feature vectors that are output from said first extracting means and the spatially-based feature vectors that are output from said second extracting means to determine first probabilities that the temporally-based feature vectors represent predetermined first prototype characters and to determine second probabilities that the spatially-based feature vectors represent predetermined second prototype characters, said processing means further including means for determining, based on a weighted combination of said first probabilities and said second probabilities, a third probability that the x-y coordinate information generated by the writer represents one of a plurality of predetermined characters.
Claim 3
Dependent Claim
346 words
Handwriting recognition apparatus, comprising: handwriting transducer means having an output providing x-y coordinate information generated by a writer while writing a character; first extracting means, having an input coupled to said output of said handwriting transducer means, for partitioning the x-y coordinate information into first partitions, and for extracting and outputting temporally-based feature vectors from first partitions of the x-y coordinate information; first character prototype storage means for storing character prototypes that are constructed in accordance with temporally-based feature vectors; first probability estimator means, having a first input coupled to an output of said first extracting means and a second input coupled to an output of said first character prototype storage means, for determining first probabilities that the temporally-based feature vectors represent character prototypes stored in said first character prototype storage means; second extracting means, having an input coupled to said output of said handwriting transducer means, for partitioning the x-y coordinate information into second partitions, and for extracting and outputting spatially-based feature vectors from the second partitions of the x-y coordinate information; second character prototype storage means for storing character prototypes that are constructed in accordance with spatially-based feature vectors; second probability estimator means, having a first input coupled to an output of said second extracting means and a second input coupled to an output of said second character prototype storage means, for determining second probabilities that the spatially-based feature vectors represent character prototypes stored in said second character prototype storage means; third probability estimator means that is responsive to an output of said first probability estimator means and to an output of said second probability estimator means for determining at least one most probable character that is represented by a weighted combination of the first probabilities and the second probabilities; and decoder means having a first input coupled to an output of said third probability estimator means and a second input coupled to an output of a language model means for identifying a most probable character that is written by the writer, said language model means providing allowable character sequences to said decoder means.
Claim 4
Dependent Claim
346 words
Handwriting recognition apparatus, comprising: handwriting transducer means having an output providing x-y coordinate information generated by a writer while writing a character; first extracting means, having an input coupled to said output of said handwriting transducer means, for partitioning the x-y coordinate information into first partitions, and for extracting and outputting temporally-based feature vectors from first partitions of the x-y coordinate information; first character prototype storage means for storing character prototypes that are constructed in accordance with temporally-based feature vectors; first probability estimator means, having a first input coupled to an output of said first extracting means and a second input coupled to an output of said first character prototype storage means, for determining first probabilities that the temporally-based feature vectors represent character prototypes stored in said first character prototype storage means; second extracting means, having an input coupled to said output of said handwriting transducer means, for partitioning the x-y coordinate information into second partitions, and for extracting and outputting spatially-based feature vectors from the second partitions of the x-y coordinate information; second character prototype storage means for storing character prototypes that are constructed in accordance with spatially-based feature vectors; second probability estimator means, having a first input coupled to an output of said second extracting means and a second input coupled to an output of said second character prototype storage means, for determining second probabilities that the spatially-based feature vectors represent character prototypes stored in said second character prototype storage means; third probability estimator means that is responsive to an output of said first probability estimator means and to an output of said second probability estimator means for determining at least one most probable character that is represented by a weighted combination of the first probabilities and the second probabilities; and decoder means having a first input coupled to an output of said third probability estimator means and a second input coupled to an output of a language model means for identifying a most probable character that is written by the writer, said language model means providing allowable character sequences to said decoder means.
Claim 5
Dependent Claim
221 words
A handwriting recognition system, comprising: a handwriting transducer generating a time-ordered data stream for representing a user's handwriting; first handwriting processor means having an input coupled to an output of said handwriting transducer for partitioning said time-ordered data stream into a plurality of first partitions, said first handwriting processor means including means for generating first feature vectors from temporal features of the first partitions of the time-ordered data stream and for estimating first probabilities that the first feature vectors represent first handwriting symbol prototypes; second handwriting processor means having an input coupled to said output of said handwriting transducer for partitioning said time-ordered data stream into a plurality of second partitions, said second handwriting processor means including means for generating second feature vectors derived from spatial features and also from selected temporal features of the second partitions of the time-ordered data stream and for estimating second probabilities that the second feature vectors represent second handwriting symbol prototypes; and means, having a first input coupled to an output of said first handwriting processor and a second input coupled to an output of said second handwriting processor, for determining a set of third probabilities based on a weighted combination of said first probabilities and said second probabilities, and for recognizing a most probable handwritten symbol in accordance with said set of third probabilities.
Claim 6
Dependent Claim
221 words
A handwriting recognition system, comprising: a handwriting transducer generating a time-ordered data stream for representing a user's handwriting; first handwriting processor means having an input coupled to an output of said handwriting transducer for partitioning said time-ordered data stream into a plurality of first partitions, said first handwriting processor means including means for generating first feature vectors from temporal features of the first partitions of the time-ordered data stream and for estimating first probabilities that the first feature vectors represent first handwriting symbol prototypes; second handwriting processor means having an input coupled to said output of said handwriting transducer for partitioning said time-ordered data stream into a plurality of second partitions, said second handwriting processor means including means for generating second feature vectors derived from spatial features and also from selected temporal features of the second partitions of the time-ordered data stream and for estimating second probabilities that the second feature vectors represent second handwriting symbol prototypes; and means, having a first input coupled to an output of said first handwriting processor and a second input coupled to an output of said second handwriting processor, for determining a set of third probabilities based on a weighted combination of said first probabilities and said second probabilities, and for recognizing a most probable handwritten symbol in accordance with said set of third probabilities.
Claim 7
Independent Claim
145 words
A method for recognizing handwriting with a handwriting recognition system, comprising the steps of: generating a time-ordered data stream for representing a user's handwriting; partitioning the time-ordered data stream into a plurality of first partitions; generating first feature vectors from temporal features of the first partitions of the time-ordered data stream; estimating first probabilities that the first feature vectors represent first handwriting symbol prototypes; partitioning the time-ordered data stream into a plurality of second partitions; generating second feature vectors derived from spatial features and also from selected temporal features of the second partitions of the time-ordered data stream; estimating second probabilities that the second feature vectors represent second handwriting symbol prototypes; determining a set of third probabilities based on a weighted combination of the first probabilities and the second probabilities; and recognizing a most probable handwritten symbol in accordance with the set of third probabilities.
Claim 8
Independent Claim
145 words
A method for recognizing handwriting with a handwriting recognition system, comprising the steps of: generating a time-ordered data stream for representing a user's handwriting; partitioning the time-ordered data stream into a plurality of first partitions; generating first feature vectors from temporal features of the first partitions of the time-ordered data stream; estimating first probabilities that the first feature vectors represent first handwriting symbol prototypes; partitioning the time-ordered data stream into a plurality of second partitions; generating second feature vectors derived from spatial features and also from selected temporal features of the second partitions of the time-ordered data stream; estimating second probabilities that the second feature vectors represent second handwriting symbol prototypes; determining a set of third probabilities based on a weighted combination of the first probabilities and the second probabilities; and recognizing a most probable handwritten symbol in accordance with the set of third probabilities.
Claim 9
Independent Claim
147 words
A method for recognizing handwriting with a handwriting recognition system, comprising the steps of: receiving, from a handwriting transducer means, a time-ordered sequence of x-y coordinate information generated by a writer; windowing the time-ordered sequence of x-y coordinate information and extracting temporally-based feature vectors from the windowed x-y coordinate information; estimating first probabilities that the extracted temporally-based feature vectors represent first handwriting symbol prototypes; blocking the time-ordered x-y coordinate information and extracting spatially-based feature vectors from the blocked x-y coordinate information; estimating second probabilities that the extracted spatially-based feature vectors represent second handwriting symbol prototypes; determining a set of third probabilities based on a weighted combination of the first probabilities and the second probabilities; and identifying a most probable character that is written by the writer in accordance with the determined set of third probabilities and in accordance with a predetermined list of allowable sequences of characters.
Claim 10
Independent Claim
147 words
A method for recognizing handwriting with a handwriting recognition system, comprising the steps of: receiving, from a handwriting transducer means, a time-ordered sequence of x-y coordinate information generated by a writer; windowing the time-ordered sequence of x-y coordinate information and extracting temporally-based feature vectors from the windowed x-y coordinate information; estimating first probabilities that the extracted temporally-based feature vectors represent first handwriting symbol prototypes; blocking the time-ordered x-y coordinate information and extracting spatially-based feature vectors from the blocked x-y coordinate information; estimating second probabilities that the extracted spatially-based feature vectors represent second handwriting symbol prototypes; determining a set of third probabilities based on a weighted combination of the first probabilities and the second probabilities; and identifying a most probable character that is written by the writer in accordance with the determined set of third probabilities and in accordance with a predetermined list of allowable sequences of characters.
References Cited (Backward Citations)
GB2190778A
Cited
US4903312A
Cited
US5058182A
Cited
DE3822671A1
Cited
US5265174A
Cited
US5005205A
Cited
WO1992005517A1
Cited
US5105468A
Cited
US5111512A
Cited
US5343537A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1999-02-12 | 1999-08-14 | 2000-02-13 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2003-02-12 | 2003-08-14 | 2004-02-13 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2007-02-12 | 2007-08-14 | 2008-02-13 | $7700.00 | Window Closed / Lapsed |