US4491960A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Handprinted symbol recognition system
Assignee: United States Department of the Navy
•
Filing: 1982-04-05
•
Granted: 1985-01-01
•
Expires: 2002-04-05
Abstract
Abstract A handprinted symbol recognition system for identifying through computer techniques free-form, unconstrained handprinting in which accuracy is decoupled from efficiency converts a thin-line (skeletonized) figure output from a preprocessing system, in the form of an array of "black" image points from a raster line sampling, into segment-oriented lists. The segment-oriented lists are filtered and data compressed to obtain a reduced more suitable subset of points approximating the symbol for further processing. The topologic features of the image are extracted from the subset. Short segments that are potential spurs are removed, the image is rotated as necessary and normalized as to size. Each segment and their interrelationships are analyzed to extract geometric features for use in conjunction with the topologic features in a logic tree decision mechanism which uses three modules: (1) pre-recognition ("trash" filter), (2) potential symbol identification, and (3) final quality assurance. The pre-recognition module separates all possible non-meaningful images from those the recognition system has a chance of identifying with minimal misclassification. The symbol identification module uses a decision tree approach in which the most reliable and "rugged" features (relative to real-world scanner data) are placed at the top of the tree. The quality assurance module asks more precise questions about the tentatively identified symbol to determine if it is indeed the particular symbol in question, i.e., the symbol meets a sufficient class membership (SCM) criterion.
Inventors
- 👤 Robert M. Brown
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-27 10:44
CPC Classifications
G; G06; G06V; G06V10/00; G06V10/20; G06V10/32; G06V10/34; G06V10/40
Claims Analysis
Total Claims: 10
Claim 1
Independent Claim
126 words
A handprinted symbol recognition system comprising: means for obtaining topologic and geometric rugged features from a skeletonized raster scan line sample image representative of a single handprinted symbol; means for filtering said topologic and geometric features by testing against given parameters to eliminate said image when said image has no characteristic common to any symbol within a target subset of symbols; means using a feature selection/decision tree approach for tentatively identifying said handprinted symbol from said target subset corresponding to said image when said image is passed by said filtering means; and means for assuring that said handprinted symbol identified by said identifying means satisfies a sufficient class membership criterion by applying more specific features to said handprinted symbol; whereby identification accuracy is decoupled from efficiency.
Claim 2
Dependent Claim
126 words
A handprinted symbol recognition system comprising: means for obtaining topologic and geometric rugged features from a skeletonized raster scan line sample image representative of a single handprinted symbol; means for filtering said topologic and geometric features by testing against given parameters to eliminate said image when said image has no characteristic common to any symbol within a target subset of symbols; means using a feature selection/decision tree approach for tentatively identifying said handprinted symbol from said target subset corresponding to said image when said image is passed by said filtering means; and means for assuring that said handprinted symbol identified by said identifying means satisfies a sufficient class membership criterion by applying more specific features to said handprinted symbol; whereby identification accuracy is decoupled from efficiency.
Claim 3
Dependent Claim
107 words
A handprinted symbol recognition system as recited in claim 1 wherein said obtaining means comprises: means for segmenting said skeletonized image to form segment-oriented lists; a topology feature generator to identify from said segment-oriented lists selected topologic features; means for orienting and normalizing said segment-oriented lists to produce a real number array for said image; and a geometric feature generator to analyze each segment of said real number array and the interrelationship between segments to calculate selected geometric features; said topologic and geometric features being defined to be rugged to noise, tolerant of variations of said handprinted symbol, and discriminatory between different ones of said handprinted symbols.
Claim 4
Dependent Claim
49 words
A handprinted symbol recognition system as recited in claim 2 wherein said filtering means comprises a pre-recognition filter which tests said topologic and geometric features against given parameters to reject said image which has no characteristic common to any recognizable handprinted symbol within a given target subset of symbols.
Claim 5
Dependent Claim
41 words
A handprinted symbol recognition system as recited in claim 3 wherein said identifying means comprises a decision tree logic which tests said topologic and geometric features for said image by means of successive binary decisions to tentatively identify said handprinted symbol.
Claim 6
Dependent Claim
55 words
A handprinted symbol recognition system as recited in claim 4 wherein said assuring means comprises a post-recognition module which tests said tentatively identified handprinted symbol against more specific ones of said topologic and geometric features to lower the possibility of a misclassification by establishing said tentatively identified handprinted symbol meets said sufficient class membership criterion.
Claim 7
Dependent Claim
56 words
A handprinted symbol recognition system as recited in claims 2 or 5 wherein said obtaining means further comprises an approximation module to act as a filter and data compressor for said segment-oriented lists by reducing the number of points to describe each segment and to filter out local variations prior to identifying said selected topologic features.
Claim 8
Dependent Claim
37 words
A handprinted symbol recognition system as recited in claim 6 wherein said obtaining means further comprises means for identifying and removing short segments that are potential spurs prior to analyzing each segment for said selected geometric features.
Claim 9
Dependent Claim
117 words
An improved handprinted symbol recognition system having geometric and topologic feature information extracted from a skeletonized input image, the improvement comprising: a pre-recognition module to separate all possible non-meaningful images from a target set of symbols based upon said geometric and topologic feature information; a symbol identification module using a decision tree logic in which the most reliable of said geometric and topologic feature information emphasizing global properties of said input image is placed at the top of said decision tree logic to select the symbol from said target set corresponding to said input image; and a quality assurance module to ask more precise questions to determine that the selected symbol meets a sufficient class membership criterion.
Claim 10
Dependent Claim
117 words
An improved handprinted symbol recognition system having geometric and topologic feature information extracted from a skeletonized input image, the improvement comprising: a pre-recognition module to separate all possible non-meaningful images from a target set of symbols based upon said geometric and topologic feature information; a symbol identification module using a decision tree logic in which the most reliable of said geometric and topologic feature information emphasizing global properties of said input image is placed at the top of said decision tree logic to select the symbol from said target set corresponding to said input image; and a quality assurance module to ask more precise questions to determine that the selected symbol meets a sufficient class membership criterion.
References Cited (Backward Citations)
US3430198A
Cited
US3636513A
Cited
US3999161A
Cited
US4040009A
Cited
US4193056A
Cited
US4259661A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1988-01-01 | 1988-07-02 | 1989-01-01 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1992-01-01 | 1992-07-02 | 1993-01-01 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 1996-01-01 | 1996-07-02 | 1997-01-01 | $7700.00 | Window Closed / Lapsed |