System and method for automated symbolic recognition including logical symbols

• US 7,236,631 B1
• Filed: 04/18/2003
• Issued: 06/26/2007
• Est. Priority Date: 04/18/2003
• Status: Expired due to Fees

• Current Assignee: Kamran Reihani
• Original Assignee: Kamran Reihani
• Inventors: Reihani, Kamran
• Primary Examiner(s): Mancuso; Joseph
• Assistant Examiner(s): DULANEY, KATHLEEN YUAN
• Application Number:
  US10/419,481
• Time in Patent Office: 1,530 Days
• Field of Search: 382/186
• US Class Current: 382/186
• CPC Class Codes: G06V 30/347 Sampling; Contour coding; Stroke extraction

Abstract

An automated symbolic recognition system and method includes the basis for the system'"'"'s feature set by identifying (a) a set of logical symbols comprising a finite class of arcpolys (lines and arcs and a point) that to the exclusion of the point, each member class has a unique (distinct) orientation, and (b) a set of subclass symbols per logical class of symbol representing a finite subclass of arcpolys (lines and arcs and a point) that to the exclusion of the point, each subclass member has a unique (distinct) extreme points size and/or depth size.

Claims

1. A method of automatically recognizing alphanumeric symbols, comprising:

(a) providing a set of logical symbols comprising a finite class of arcpolys, wherein the class of arcpolys comprises arcs, lines and points;

(b) providing a set of subclass symbols per logical symbol representing a finite subclass of arcpolys, each subclass member having a unique extreme points size and/or depth size;

(c) providing a digital representation of a set of unrecognized alphanumeric symbols;

(d) automatically classifying the arcpolys of each unrecognized alphanumeric symbol into one or more pairs of logical and subclass symbols such that each unrecognized alphanumeric symbol can be recognized; and

(e) providing a set of topological relationships for each pair of arcpolys each comprising a one-of-nine topological connection code and separation, and wherein topological relational features are also automatically classified.

2. The method of claim 1, wherein each pair of logical and subclass symbols is characterized by a set of primary features including at least the following group of features:

(a) extreme points direction, (b) extreme points size, (c) depth direction, and (d) depth size.

3. The method of claim 2, wherein the primary features further include at least one of:

(e) extension and (f) motion clockwise.

14. The method of claim 1, wherein the automatically classifying alphanumerics comprises:

computing a set of arcpolys for each unrecognized alphanumeric symbol;

computing a feature set based on one or more sets of primary features for each unrecognized alphanumeric symbol; and

computing a symbolic representation for each unrecognized alphanumeric symbol so that the unrecognized alphanumeric symbols can be recognized.

23. The method of claim 1, wherein each arc and line member of the logical symbol set has a unique orientation.

4. A method of automatically recognizing alphanumeric symbols, comprising:

receiving a digital representation of a set of unrecognized alphanumeric symbols;

computing a set of arcpolys for each unrecognized alphanumeric symbol, wherein the set of arcpolys is a subset of the class of arcpolys which comprises arcs, lines and points;

computing a feature set based on one or more sets of primary features for each unrecognized alphanumeric symbol;

computing a symbolic representation for each unrecognized alphanumeric symbol so that the unrecognized alphanumeric symbols can be recognized; and

providing a set of topological relationships for each pair of arcpolys each comprising a one-of-nine topological connection code and separation, and wherein topological relational features are also automatically classified.

5. The method of claim 4, wherein the computing a feature set comprises:

normalizing the primary features sizes of the arcpolys, wherein primary features represent shape and orientation of the arcpoly;

grouping the normalized primary features of the arcpolys into each unrecognized alphanumeric symbol;

determining topological relationships for each pair of arcpolys; and

grouping the topological relationships of arcpoly pair(s) into each unrecognized alphanumeric symbol.

6. The method of claim 5, wherein the computing a feature set additionally comprises criteria-based region growing and splitting method to achieve arcpolys that better conform to a plurality of alphanumeric symbol models.

7. The method of claim 4, wherein the computing a symbolic representation comprises for each arcpoly of each polyline of each unrecognized alphanumeric symbol, computing primary relational features from the unrecognized alphanumeric symbol'"'"'s feature set per arcpoly.

8. The method of claim 4, wherein the computing symbolic representations for each arcpoly of each polyline of each unrecognized symbol comprises computing one or more pairs of logical and subclass symbols.

9. The method of claim 4, additionally comprising:

for each arcpoly of each polyline of each unrecognized alphanumeric symbol;

accessing a stored candidate alphanumeric list;

computing a reduced alphanumeric candidate list from a database of alphanumeric symbol models; and

computing one or more pairs of logical and subclass symbols and secondary relational features for each candidate of the reduced alphanumeric candidate list.

10. The method of claim 9, wherein the computing pairs of logical and subclass symbols and secondary relational features comprises:

for each arcpoly of each polyline of each unrecognized alphanumeric symbol;

computing secondary relational features comprising topological variance and arcpoly dissimilarity value representing inferred confidence level, as each unrecognized alphanumeric symbol is paired with each candidate alphanumeric symbol from the database; and

discarding candidate alphanumeric symbols to reduce the candidate alphanumeric list when selective feature values of the secondary relational feature set based on each candidate alphanumeric symbol exceed preset criteria-based thresholds.

11. The method of claim 10, wherein the secondary relational features are used to derive the alphanumeric dissimilarity value based on each candidate alphanumeric symbol.

22. The method of claim 4, wherein the computing a symbolic representation for each unrecognized alphanumeric comprises computing one or more pairs of logical and subclass symbols per arcpoly.

12. A computerized database stored in a digital storage medium and accessed by a computing device to automatically recognize alphanumeric symbols, comprising:

a set of logical symbols comprising a finite class of arcpolys, wherein the class of arcpolys comprises arcs, lines and points;

a set of subclass symbols per logical symbol representing a finite subclass of arcpolys, each subclass member having a unique extreme points size and/or depth size;

wherein each subclass symbol is characterized by a group of features including motion clockwise; and

wherein a set of topological relationships for each pair of arcpolys each comprises a one-of-nine topological connection code and separation, and wherein topological relational features are also automatically classified.

13. The method of claim 12, additionally comprising a set of variances comprising a set of arcpoly unit variance types, a set of counterpart dissimilarity values and a unit topological dissimilarity value per pair of arcpolys.

24. The computerized database of claim 12, wherein each arc and line member of the logical symbol set has a unique orientation.

25. The computerized database of claim 12, wherein each subclass symbol is additionally characterized by the feature of extension.

15. A system for automatically recognizing alphanumeric symbols, comprising:

means for providing a set of logical symbols comprising a finite class of arcpolys, wherein the class of arcpolys comprises arcs, lines and points;

means for providing a set of subclass symbols per logical symbol representing a finite subclass of arcpolys, each subclass member having a unique extreme points size and/or depth size;

means for providing a digital representation of a set of unrecognized alphanumeric symbols;

means for automatically classifying the arcpolys of each unrecognized alphanumeric symbol into one or more pairs of logical and subclass symbols such that each unrecognized alphanumeric symbol can be recognized; and

means for providing a set of topological relationships for each pair of arcpolys each comprising a one-of-nine topological connection code and separation, and wherein topological relational features are also automatically classified.

16. The method of claim 15, wherein each pair of logical and subclass symbols is primarily characterized by a set of primary features including at least the following group of features:

(a) extreme points direction, (b) extreme points size, (c) depth direction, and (d) depth size.

17. The method of claim 16, wherein the primary features further include at least one of:

(e) extension and (f) motion clockwise.

26. The system of claim 15, wherein each arc and line member of the logical symbol set has a unique orientation.

18. A system for automatically recognizing alphanumeric symbols, comprising:

means for receiving a digital representation of a set of unrecognized alphanumeric symbols;

means for computing a set of arcpolys for each unrecognized alphanumeric symbol, wherein the set of arcpolys is a subset of the class of arcpolys which comprises arcs, lines and points;

means for computing a feature set based on one or more sets of primary features for each unrecognized alphanumeric symbol;

means for computing a symbolic representation for each unrecognized alphanumeric symbol so that the unrecognized alphanumeric symbols can be recognized; and

means for providing a set of topological relationships for each pair of arcpolys each comprising a one-of-nine topological connection code and separation, and wherein topological relational features are also automatically classified.

19. The system of claim 18, wherein the computing a feature set comprises:

means for normalizing the primary features sizes of the arcpolys;

means for grouping the normalized primary features of the arcpolys into each unrecognized alphanumeric symbol;

means for determining topological relationships for each pair of arcpolys; and

means for grouping the topological relationships of arcpoly pair(s) into each unrecognized alphanumeric symbol.

20. The system of claim 18, wherein the means for computing symbolic representations for each arcpoly of each polyline of each unrecognized symbol comprises means for computing one or more pairs of logical and subclass symbols.

21. A system for automatically recognizing alphanumeric symbols, comprising:

a first data storage having a set of logical symbols comprising a finite class of arcpolys, wherein the class of arcpolys comprises arcs, lines and points;

a second data storage having a set of subclass symbols per logical symbol representing a finite subclass of arcpolys, each subclass member having a unique extreme points size and/or depth size;

a third data storage having a digital representation of a set of unrecognized alphanumeric symbols;

a fourth data storage having structural and topological multi-type variances and counterpart dissimilarity values comprising the following;

(a) unit change extreme points size, (b) unit change depth size, (c) extra image structure size, (d) missing stored symbol size, (e) unit change single extreme point position, (f) unit change single extreme point extension, and (g) unit rotation, and

program code accessing the data storages and adapted to automatically classify the arcpolys of each unrecognized alphanumeric symbol into one or more pairs of logical and subclass symbols such that each unrecognized alphanumeric symbol can be recognized.

27. The system of claim 21, wherein each arc and line member of the logical symbol set has a unique orientation.