Using edges and corners for character input

• US 7,729,542 B2
• Filed: 03/29/2004
• Issued: 06/01/2010
• Est. Priority Date: 04/04/2003
• Status: Active Grant

• Current Assignee: Carnegie Mellon University
• Original Assignee: Carnegie Mellon University
• Inventors: Myers, Brad A., Wobbrock, Jacob O.
• Primary Examiner(s): PERUNGAVOOR, SATHYANARAYA V
• Application Number:
  US10/811,761
• Publication Number:
  US 20040196256A1
• Time in Patent Office: 2,255 Days
• Field of Search: 382/187
• US Class Current: 382/187
• CPC Class Codes: G04G 21/08 Touch switches specially adapted for time-pieces
• CPC Class Codes: G06F 1/1626 with a single-body enclosure integrating a flat display, e.g. Personal Digital Assistants [PDAs]
• CPC Class Codes: G06F 1/163 Wearable computers, e.g. on a belt
• CPC Class Codes: G06F 1/169 the I/O peripheral being an integrated pointing device, e.g. trackball in the palm rest area, mini-joystick integrated between keyboard keys, touch pads or touch stripes G06F1/1643 takes precedence; constructional details of pointing devices G06F3/033
• CPC Class Codes: G06F 3/0233 Character input methods
• CPC Class Codes: G06F 3/04883 for inputting data by handwriting, e.g. gesture or text
• CPC Class Codes: G06F 3/04886 by partitioning the display area of the touch-screen or the surface of the digitising tablet into independently controllable areas, e.g. virtual keyboards or menus

Abstract

A new unistroke text entry method for handheld or wearable devices is designed to provide high accuracy and stability of motion. The user makes characters by traversing the edges and diagonals of a geometric pattern, e.g. a square, imposed over the usual text input area. Gesture recognition is accomplished not through pattern recognition but through the sequence of corners that are hit. This means that the full stroke path is unimportant and the recognition is highly deterministic, enabling better accuracy than other gestural alphabets. This input technique works well using a template with a square hole placed over a touch-sensitive surface, such as on a Personal Digital Assistant (PDA), and with a square boundary surrounding a joystick, which might be used on a cell-phone or game controller. Another feature of the input technique is that capital letters are made by ending the stroke in a particular corner, rather than through a mode change as in other gestural input techniques. Because of the rules governing abstracts, this abstract should not be used to construe the claims.

Claims

1. A character recognition method, comprising:

using a processor to perform the following method;

queuing continuous path data made by traversing edges and diagonals and into corners within a confining geometric shape constraining an input device to less than the input device'"'"'s total area;

searching the queued continuous path data to generate a sequence of corner hits, where a corner hit corresponds to a corner defined by said geometric shape;

identifying a character based on said sequence of corner hits independently of the remainder of the continuous path data;

wherein each of said sequences of corner hits defines a single stroke, and wherein each single stroke is representative of one of a letter, number, punctuation or mode; and

identifying a letter character as being upper case when said stroke representative of said character ends in a common predetermined corner and lower case when said stroke does not end in said common predetermined corner.

2. The method of claim 1 wherein said input device is a touch sensitive surface, said method additionally comprising detecting loss of contact with the touch sensitive surface, said loss of contact indicating the end of a stroke.

3. The method of claim 1 additionally comprising detecting the actuation of a switch, said actuation indicating the end of a stroke.

4. The method of claim 1 wherein said input device is a joystick, said method additionally comprising detecting lack of movement of the joystick for a predetermined period of time, said lack of movement indicating the end of a stroke.

5. The method of claim 4 wherein said detecting lack of movement includes detecting the joystick at two identical positions within said predetermined period of time.

6. The method of claim 5 wherein said positions correspond to a center point.

7. The method of claim 1 wherein said identifying a character is comprised of comparing the determined sequence of corner hits to data representative of a plurality of stored sequences of corner hits, selecting one of the stored sequences of corner hits based on said comparing, and outputting a character linked to said selected one of said stored sequences of corner hits.

8. The method of claim 7 wherein said comparing includes comparing the determined sequence of corner hits to a library of stored sequences of corner hits which is representational of a printed alphabet.

9. The method of claim 7 additionally comprising changing the stored sequences of corner hits that are linked to a character.

10. The method of claim 9 wherein said changing includes providing one example of a sequence of corner hits and the character to which that sequence is to be linked.

11. The method of claim 1 wherein said corner hits include corner area hits, said method additionally comprising varying the size of the corner areas while said continuous path data is being queued.

12. The method of claim 11 wherein said varying the size includes decreasing the size of only certain corner areas.

13. The method of claim 11 wherein said varying the size includes decreasing the size of certain corner areas more than the size of other corner areas.

14. The method of claim 1 wherein said corner hits include corner area hits, said method additionally comprising varying the shape of the corner areas while said continuous path data is being queued.

15. A letter character recognition method, comprising:

using a processor to perform the following method;

queuing continuous path data made by traversing edges and diagonals and into corners within a confining geometric shape constraining an input device to less than the input device'"'"'s total area;

searching the queued continuous path data to generate a sequence of corner hits within a single unistroke, where a corner hit corresponds to a corner defined by said geometric shape;

identifying a letter character based on said sequence of corner hits independently of the path therebetween; and

identifying said letter character as being upper case when said single unistroke defining the character ends in a common predetermined corner and lower case when said unistroke does not end in the common predetermined corner;

wherein each of said sequences of corner hits defines a single stroke, and wherein each single stroke is representative of one of a letter, number, punctuation or mode.

16. The method of claim 15 additionally comprising detecting loss of contact with a touch sensitive surface, said loss of contact indicating the end of the unistroke.

17. The method of claim 15 additionally comprising detecting the actuation of a switch, said actuation indicating the end of the unistroke.

18. The method of claim 15 additionally comprising detecting lack of movement of a joystick for a predetermined period of time, said lack of movement indicating the end of the unistroke.

19. The method of claim 18 wherein said detecting lack of movement includes detecting the joystick at two identical positions within said predetermined period of time.

20. The method of claim 19 wherein said positions correspond to a center point.

21. The method of claim 15 wherein said identifying a letter character is comprised of comparing the determined sequence of corner hits to data representative of a plurality of stored sequences of corner hits, selecting one of the stored sequences of corner hits based on said comparing, and outputting the letter character linked to said selected one of said stored sequences of corner hits.

22. The method of claim 21 wherein said comparing includes comparing the determined sequence of corner hits to a library of stored sequences of corner hits which is representational of a printed alphabet.

23. The method of claim 21 additionally comprising changing the stored sequences of corner hits that are linked to a letter character.

24. The method of claim 23 wherein said changing includes providing one example of a sequence of corner hits and the letter character to which that sequence is to be linked.

25. The method of claim 15 wherein said corner hits include corner area hits, said method additionally comprising varying the size of the corner areas while said single unistroke is created.

26. The method of claim 25 wherein said varying the size includes decreasing the size of only certain corner areas.

27. The method of claim 25 wherein said varying the size includes decreasing the size of certain corner areas more than the size of other corner areas.

28. The method of claim 15 wherein said corner hits include corner area hits, said method additionally comprising varying the shape of the corner areas while said single unistroke is created.

29. A method of generating a stroke, comprising:

using a processor to perform the following method;

queuing continuous path data made by traversing edges and diagonals and into corners within a confining geometric shape constraining an input device to less than the input device'"'"'s total area;

searching the queued continuous path data to generate a sequence of corner hits independently of the remainder of the continuous path data, with each corner hit in said sequence of corner hits corresponding to a corner defined by said geometric shape;

receiving information indicative of the end of each stroke;

wherein each of said sequences of corner hits defines a single stroke, and wherein each single stroke is representative of one of a letter, number, punctuation or mode; and

identifying a letter character as being upper case when said stroke representative of said character ends in a common predetermined corner and lower case when said stroke does not end in said common predetermined corner.

30. The method of claim 29 wherein said input device is a touch sensitive surface, and wherein said information includes information generated by lifting an object out of contact with the touch sensitive surface.

31. The method of claim 29 wherein said information includes information generated by activating a switch.

32. The method of claim 29 wherein said input device is a joystick, and wherein said information includes information generated by returning the joystick to a predetermined position for a predetermined period of time.

33. A computer readable memory carrying software which, when executed, performs a method comprising:

queuing continuous path data made by traversing edges and diagonals and into corners within a confining geometric shape constraining an input device to less than the input device'"'"'s total area;

searching the queued continuous path data to generate a sequence of corner hits, where a corner hit corresponds to a corner defined by said geometric shape;

identifying a character based on said sequence of corner hits independently of the remainder of the continuous path data;

wherein each of said sequences of corner hits defines a single stroke, and wherein each single stroke is representative of one of a letter, number, punctuation or mode; and

identifying a letter character as being upper case when said stroke representative of said character ends in a common predetermined corner and lower case when said stroke does not end in said common predetermined corner.

34. The memory of claim 33 wherein said input device is a touch sensitive surface, said method additionally comprising detecting loss of contact with the touch sensitive surface, said loss of contact indicating the end of a stroke.

35. The memory of claim 33 additionally comprising detecting the actuation of a switch, said actuation indicating the end of a stroke.

36. The memory of claim 33 wherein said input device is a joystick, said method additionally comprising detecting lack of movement of the joystick for a predetermined period of time, said lack of movement indicating the end of a stroke.

37. The memory of claim 36 wherein said detecting lack of movement includes detecting the joystick at two identical positions within said predetermined period of time.

38. The memory of claim 37 wherein said positions correspond to a center point.

39. The memory of claim 33 wherein said identifying a character is comprised of comparing the determined sequence of corner hits to data representative of a plurality of stored sequences of corner hits, selecting one of the stored sequences of corner hits based on said comparing, and outputting a character linked to said selected one of said stored sequences of corner hits.

40. The memory of claim 39 wherein said comparing includes comparing the determined sequence of corner hits to a library of stored sequences of corner hits which is representational of a printed alphabet.

41. The memory of claim 39 additionally comprising changing the stored sequences of corner hits that are linked to a character.

42. The memory of claim 41 wherein said changing includes providing one example of a sequence of corner hits and the character to which that sequence is to be linked.

43. The memory of claim 33 wherein said corner hits include corner area hits, said method additionally comprising varying the size of the corner areas while said continuous path data is being queued.

44. The memory of claim 43 wherein said varying the size includes decreasing the size of only certain corner areas.

45. The memory of claim 43 wherein said varying the size includes decreasing the size of certain corner areas more than the size of other corner areas.

46. The memory of claim 33 wherein said corner hits include corner area hits, said method additionally comprising varying the shape of the corner areas while said continuous path data is being queued.

47. A computer readable memory carrying software which, when executed, performs a method, comprising:

queuing continuous path data made by traversing edges and diagonals and into corners within a confining geometric shape constraining an input device to less than the input device'"'"'s total area;

searching the queued continuous path data to generate a sequence of corner hits within a single unistroke, where a corner hit corresponds to a corner defined by said geometric shape;

identifying a letter character based on said sequence of corner hits independently of the path therebetween; and

identifying said letter character as being upper case when said single unistroke defining the character ends in a common predetermined corner and lower case when said unistroke does not end in the common predetermined corner;

wherein each of said sequences of corner hits defines a single stroke, and wherein each single stroke is representative of one of a letter, number, punctuation or mode.

48. The memory of claim 47 additionally comprising detecting loss of contact with a touch sensitive surface, said loss of contact indicating the end of the unistroke.

49. The memory of claim 47 additionally comprising detecting the actuation of a switch, said actuation indicating the end of the unistroke.

50. The memory of claim 47 additionally comprising detecting lack of movement of a joystick for a predetermined period of time, said lack of movement indicating the end of the unistroke.

51. The memory of claim 50 wherein said detecting lack of movement includes detecting the joystick at two identical positions within said predetermined period of time.

52. The memory of claim 51 wherein said positions correspond to a center point.

53. The memory of claim 47 wherein said identifying a letter character is comprised of comparing the determined sequence of corner hits to data representative of a plurality of stored sequences of corner hits, selecting one of the stored sequences of corner hits based on said comparing, and outputting the letter character linked to said selected one of said stored sequences of corner hits.

54. The memory of claim 53 wherein said comparing includes comparing the determined sequence of corner hits to a library of stored sequences of corner hits which is representational of a printed alphabet.

55. The memory of claim 53 additionally comprising changing the stored sequences of corner hits that are linked to a letter character.

56. The memory of claim 55 wherein said changing includes providing one example of a sequence of corner hits and the letter character to which that sequence is to be linked.

57. The memory of claim 47 wherein said corner hits include corner area hits, said method additionally comprising varying the size of the corner areas while said single unistroke is created.

58. The memory of claim 57 wherein said varying the size includes decreasing the size of only certain corner areas.

59. The memory of claim 57 wherein said varying the size includes decreasing the size of certain corner areas more than the size of other corner areas.

60. The memory of claim 47 wherein said corner hits include corner area hits, said method additionally comprising varying the shape of the corner areas while said single unistroke is created.