IP Library Granted Patent US 8,351,700
Granted Patent B2
US 8,351,700 · App. 12/756,970 · Granted Jan 8, 2013

Variable glyph system and method

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 8,351,700
App. No.
12/756,970
Granted
Jan 8, 2013
Kind
B2
Abstract

Using methods, computer-readable storage media, and apparatuses for computer-implemented processing, an image of handwritten text may be segmented into a disjoint component image corresponding to individual glyphs connected by ligatures. The disjoint component image is skeletonized into a grid graph, and a connected path traversing the ligatures is determined. The disjoint component image is segmented into non-overlapping segments based on connected graphs corresponding to edges in the disjoint component image, where sets of adjacent non-overlapping segments correspond to the individual glyphs. Glyph geometry may be varied by obtaining an ensemble of glyph representations, each characterized by measurable geometric glyph properties. For each geometric glyph property, target values are obtained from ensemble-wide statistical distribution functions, and the target values are used to transform a base glyph representation into a target glyph representation.

Claims (47)

1. A computer-implemented method for processing an image of handwritten text, the method comprising:

segmenting, by a computing device, the image of handwritten text into at least one disjoint component image corresponding to a plurality of individual glyphs connected by a plurality of ligatures;

skeletonizing, by said computing device, said disjoint component image into a grid graph comprising a plurality of nodes with at least three paths;

determining, by said computing device according to said grid graph, a connected path traversing said plurality of ligatures and comprising a subset of said plurality of nodes with at least three paths;

generating, by said computing device, a plurality of connected graphs corresponding to a plurality of edges in said disjoint component image;

segmenting, by said computing device, said disjoint component image into a plurality of non-overlapping segments according to said connected path and said plurality of connected graphs; and

determining, by said computing device, a set of one or more adjacent non-overlapping segments corresponding to each of said plurality of individual glyphs.

2. The method of claim 1 , further comprising classifying said plurality of nodes with at least three paths according to at least one of a node-type property and a node-location property.

3. The method of claim 1 , further comprising classifying said plurality of connected graphs according to an exterior/interior property.

4. The method of claim 1 , wherein sectioning said disjoint component image into said plurality of non-overlapping segments comprises:

identifying a suitable dividing node within said connected path according to at least one suitability criterion; and

segmenting said disjoint component image into two of said plurality of non-overlapping segments at said suitable dividing node.

5. The method of claim 4 , wherein identifying said suitable dividing node according to at least one suitability criterion comprises determining that a horizontal distance between said suitable dividing node and a previously-identified suitable dividing node meets or exceeds a minimum horizontal-segment-size threshold.

6. The method of claim 4 , wherein identifying said suitable dividing node according to at least one suitability criterion comprises determining that a dividing line between said two of said plurality of non-overlapping segments extends between an upper coast and a lower coast of said disjoint component image.

7. The method of claim 1 , wherein determining a set of one or more adjacent non-overlapping segments corresponding to an individual glyph comprises minimizing a distance metric between a candidate image and a plurality of glyph exemplars, said candidate image comprising said one or more adjacent non-overlapping segments.

8. The method of claim 1 , further comprising storing at least one determined set of one or more adjacent non-overlapping segments as a member of an ensemble of representations of the corresponding individual glyph.

9. The method of claim 8 , further comprising:

for each of a plurality of geometric glyph properties:

calculating a measure of the current geometric glyph property for said member of said ensemble;

defining a statistical distribution function of the current geometric glyph property for said ensemble; and

obtaining a target value for the current geometric glyph property according to the current statistical distribution function; and

transforming a base representation of the corresponding individual glyph into a target representation of the corresponding individual glyph according to the target values obtained for each of said plurality of geometric glyph properties.

10. A computer-readable storage medium having stored thereon instructions that, when executed by a processor, perform the method of claim 1 .

11. An apparatus comprising a processor and a memory, the memory having stored thereon instructions that, when executed by the processor, perform the method of claim 1 .

12. A computer-implemented method for varying glyph geometry, the method comprising:

obtaining, by a computing device, an ensemble of representations of a glyph;

for each of a plurality of geometric glyph properties:

calculating, by said computing device, a measure of the current geometric glyph property for each member of said ensemble of representations of said glyph;

defining, by said computing device, a statistical distribution function of the current geometric glyph property for said ensemble of representations of said glyph; and

obtaining, by said computing device, a target value for the current geometric glyph property according to the current statistical distribution function;

obtaining, by said computing device, a base representation of said glyph; and

transforming, by said computing device, said base representation of said glyph into a target representation of said glyph according to the target values obtained for each of said plurality of geometric glyph properties.

13. The method of claim 12 , wherein said plurality of geometric glyph properties includes at least two of a glyph position-above-a-baseline property, a glyph size property, and a glyph orientation property.

14. The method of claim 12 , wherein said plurality of geometric glyph properties includes a glyph size property, and wherein calculating a measure of said glyph size property comprises calculating a first and a second principal moment of inertia of a member of said ensemble of representations of said glyph.

15. The method of claim 14 , wherein said plurality of geometric glyph properties includes a glyph orientation property, and wherein calculating a measure of said glyph orientation property comprises calculating an angle between said first principal moment of inertia and a baseline.

16. The method of claim 12 , wherein said plurality of geometric glyph properties includes a glyph position-above-a-baseline property, and wherein calculating a measure of said glyph position-above-a-baseline property comprises summing a plurality of values corresponding to a plurality of vertical coordinates, relative to a baseline, of a plurality of points that comprise a member of said ensemble of representations of said glyph.

17. The method of claim 12 , wherein obtaining said base representation of said glyph comprises selecting one from said ensemble of representations of said glyph.

18. The method of claim 12 , wherein obtaining said ensemble of representations of said glyph comprises:

segmenting an image of handwritten text into at least one disjoint component image corresponding to a plurality of individual glyphs connected by a plurality of ligatures;

sectioning said disjoint component image into a plurality of non-overlapping segments; and

determining a set of one or more adjacent non-overlapping segments corresponding to each of said plurality of individual glyphs.

19. The method of claim 18 , wherein sectioning said disjoint component image into said plurality of non-overlapping segments comprises:

skeletonizing said disjoint component image into a grid graph comprising a plurality of nodes with at least three paths;

determining, according to said grid graph, a connected path traversing said plurality of ligatures and comprising a subset of said plurality of nodes with at least three paths; and

generating, by said computing device, a plurality of connected graphs corresponding to a plurality of edges in said disjoint component image.

20. A computer-readable storage medium having stored thereon instructions that, when executed by a processor, perform the method of claim 12 .

21. An apparatus comprising a processor and a memory, the memory having stored thereon instructions that, when executed by the processor, perform the method of claim 12 .

Assignments (6)
RELEASE OF SECURITY INTEREST Recorded Jul 2, 2025
From: AETHER FINANCIAL SERVICES, AS SECURITY AGENT
To: PLANETART, LLC
Reel/Frame 071801/0212 →
SECURITY INTEREST Recorded Apr 3, 2024
From: PLANETART, LLC
To: AETHER FINANCIAL SERVICES, AS SECURITY AGENT
Reel/Frame 066988/0768 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 20, 2021
From: MAILING SERVICES OF PITTSBURGH, INC.
To: PLANETART, LLC
Reel/Frame 055979/0420 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 23, 2018
From: GRACIOUS ELOISE, INC.
To: MAILING SERVICES OF PITTSBURGH INC.
Reel/Frame 047283/0270 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 8, 2018
From: GRACIOUS ELOISE, LLC
To: GRACIOUS ELOISE, INC.
Reel/Frame 045747/0473 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 13, 2011
From: D'AGOSTINO, ELOISE BUNE; D'AGOSTINO, MICHAEL BENNETT; MINOR, BRYAN MICHAEL; FRAJKA, TAMAS; PETTIGREW, MICHEL FRANCOIS
To: GRACIOUS ELOISE, LLC
Reel/Frame 026587/0183 →