System for recognizing various input data types

• US 5,781,663 A
• Filed: 06/27/1995
• Issued: 07/14/1998
• Est. Priority Date: 06/30/1994
• Status: Expired due to Term

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Yoshii, Hiroto, Arai, Tsunekazu, Takasu, Eiji, Sakaguchi, Katsuhiko
• Primary Examiner(s): SHALWALA, BIPIN H
• Application Number:
  US08/495,102
• Time in Patent Office: 1,113 Days
• Field of Search: 382/181, 382/182, 382/184, 382/185, 382/186, 382/187, 382/189, 382/228, 235/462, 395/12, 345/173
• US Class Current: 382/189
• CPC Class Codes: G06V 30/242 Division of the character sequences into groups prior to recognition; Selection of dictionaries

Abstract

An image recognition apparatus recognizes an input locus as a command, a figure and a character, respectively in gesture recognition mode, figure recognition mode and character recognition mode. Regarding each recognized result, similarity between the input image and the recognized shape is obtained. The similarities are compared with each other, and if the difference between the similarities is less than a predetermined value, the recognized results are displayed for selection by an operator. Then, selected one of the displayed shapes is determined as the final recognition result. This enables correct locus input even if there is a possibility that the locus is recognized, in different recognition modes, as similar shapes of different functions.

Claims

1. An image recognition apparatus for recognizing an input locus, comprising:

locus input means for inputting a locus;

recognition means for recognizing an input image constituted by the locus inputted by said locus input means in a plurality of recognition modes and for obtaining first candidates of a recognition result in each of the plurality of recognition modes as a plurality of functions;

determination means for determining similarities of the first candidates so as to obtain second candidates of a recognition result, the second candidates having similar shapes to one another; and

selection means for selecting one of the second candidates from the plurality of functions as the recognition result.

2. The image recognition apparatus according to claim 1, wherein said recognition means has a plurality of recognition modes corresponding to types of an input image and performs recognition in each mode, and wherein said determination means obtains the second candidates by determining similarities in different modes of the plurality of recognition modes so that second candidates having different functions from one another are obtained.

3. The image recognition apparatus according to claim 2, wherein the plurality of recognition modes include a character recognition mode for recognizing an input image as a character and a command recognition mode for recognizing an input image as a command to edit characters of an input image.

4. The image recognition apparatus according to claim 3, further comprising memory means for storing correspondence among the plurality of functions, wherein said recognition means first performs recognition in the command recognition mode, and if the recognition is successful, said selection means refers to the correspondence in said memory means to select one shape.

5. The image recognition apparatus according to claim 2, further comprising priority input means for inputting priorities of the respective recognition modes, wherein said determination means obtains the second candidates by using the priorities.

6. The image recognition apparatus according to claim 2, wherein said recognition means adds a predetermined order to the respective recognition modes, and if recognition in one mode is unsuccessful, performs recognition in another recognition mode, in accordance with the order.

7. The image recognition apparatus according to claim 6, further comprising memory means for storing correspondence among a plurality of functions recognized in the respective recognition modes, wherein said recognition means first performs the recognition in the input image in a command recognition mode, and if the recognition is successful, said determination means refers to the correspondence in said memory means to obtain the second candidates.

8. The image recognition apparatus according to claim 6, wherein if the recognition of input image is successful in one recognition mode, said selection means selects a recognized shape from the recognition as a recognition result.

9. The image recognition apparatus according to claim 2, further comprising display means for displaying an image, wherein said display means displays the second candidates for selection of one of the second candidates.

10. The image recognition apparatus according to claim 9, further comprising reference input means for inputting a reference value of similarity, wherein selection means displays the functions in accordance with the reference value.

11. The image recognition apparatus according to claim 1, wherein said recognition means obtains similarities between the input image and shapes corresponding to the recognized candidates, and wherein said selection means selects one of the candidates based on the similarities.

12. The image recognition apparatus according to claim 11, wherein if a difference between similarities between the input image and shapes corresponding to the first candidates is less than a predetermined value, said selection means displays the shapes for selection of one shape.

13. The image recognition apparatus according to claim 12, further comprising input means for inputting the predetermined value.

14. The image recognition apparatus according to claim 11, further comprising priority input means for inputting priorities, wherein said determination means performs weighting on the similarities in accordance with the priorities.

15. An image recognition apparatus comprising:

locus input means for inputting a locus;

recognition means for recognizing an input image constituted by the locus inputted by said locus input means in a plurality of recognition modes and for obtaining first candidates of a recognition result in each of the plurality of recognition modes;

judgment means for judging similarities in shapes among the first candidates obtained in a plurality of recognition modes

display means for displaying, as second candidates, the shapes of the first candidates which are judged as having similar shares; and

selection means for selecting one, as the recognition result, from the second candidates displayed by said display means, in accordance with designation of a user.

16. An image processing method for recognizing an input locus, comprising:

a locus input step of inputting a locus;

a recognition step of recognizing an input image as shapes having different function using a plurality of recognition modes, and of obtaining first candidates of a recognition result in each of the pluarity of recognition modes;

a judgemnet step of judging similarities in the shapes of the first candidates obtained in the plurality of recognition modes in said recognition step;

a display step for displaying , as second candidates, the shape of the first candidates which are judged as having similar shapes; and

a selection step of selecting one, as the recognition result, from the second candidates displayed by the display means, in accordance with designation of a user.

17. An image processing method according to claim 16, wherein in said recognition step, the plurality of recognition modes include a character recognition mode for recognizing an input image as a character and a command recognition mode for recognizing an input image as a command to edit input characters.

18. The image processing method according to claim 16, wherein said recognition step includes a step of judging whether or not recognition is successful.

19. The image processing method according to claim 16, wherein in said recognition step, similarities between an input image and the candidate shapes are obtained.

20. The image processing method according to claim 16, further comprising a step of inputting priorities, wherein in said judgment step, the similarities are judged using the input priorities.

21. A data input apparatus for encoding an input locus and inputting code data, comprising:

recognition means for recognizing an image by (1) a locus, (2) recognizing an input image as shares having different functions using a plurality of recognition modes, (3) obtaining first candidates of a recognition result in each of the plurality of recognition modes, (4) judging similarities in the shaves of the first candidates obtained in the plurality of recognition modes, (5) displaying, as second candidates, the shape of the first candidates which are judging as having similar shares, and (6) selecting one, as the recognition result, from the second candidates displayed by the display means, in accordance with designation of a user; and

input means for inputting a code corresponding to a shape obtained as a recognition result from recognition by said recognition means.

22. An image recognition method for recognizing an input locus, comprising steps of:

inputting a locus;

recognizing an input image constituted by the input locus in a plurality of recognition modes and obtaining first candidates of a recognition result in each of the plurality of recognition modes as a plurality of functions;

determining similarities of the first candidates so as to obtain second candidates of a recognition result, the second candidates having similar shapes to one another; and

selecting one from the second candidates from the plurality of functions as the recognition result.

23. A method according to claim 22, wherein an input image is recognized in a plurality of recognition modes corresponding to types of an input image, and wherein the second candidates are obtained by determining similarities in different modes of the plurality of recognition modes so that second candidates having different functions from one another are obtained.

24. A method according to claim 23, wherein the plurality of recognition modes include a character recognition mode for recognizing an input image as a character and a command recognition mode for recognizing an input image as a command to edit characters of an input image.

25. A method according to claim 24, further comprising the step of storing similarities of the first candidates, wherein recognition is performed in the command recognition mode first, and if the recognition is successful, a corresponding similarity is referred to in said selecting step.

26. A method according to claim 23, further comprising a priority inputting step of inputting priorities of the respective recognition modes, wherein the second candidates are obtained in said determining step by using the priorities.

27. A method according to claim 23, wherein a predetermined order is added to the respective recognition modes in said recognizing step, and if recognition in one mode is unsuccessful, recognition in another recognition mode is performed, in accordance with the order.

28. A method according to claim 27, further comprising the step of storing similarities of the first candidates, wherein recognition is performed in the command recognition mode first, and if the recognition is successful, a corresponding similarity is referred to in said selecting step.

29. A method according to claim 27, wherein if the recognition of the input image is successful in one recognition mode, a recognized shape is selected from the recognition as a recognition result.

30. A method according to claim 23, further comprising a displaying step of displaying an image, wherein the first candidates which are recognized in the respective modes and have their similarities determined in said determination step are displayed in said selecting step for selection of one shape.

31. A method according to claim 30, further comprising reference inputting step of inputting a reference value of similarity, wherein the second candidates are displayed in accordance with the reference value in the selecting step.

32. A method according to claim 22, wherein similarities between the input image and the first candidates are obtained in said recognizing step, and wherein a function based on one of the similarities is selected in said selecting step.

33. A method according to claim 32, wherein if the difference between similarities of the first candidates is less than a predetermined value, the first candidates are displayed for selection in said selection step.

34. A method according to claim 33, further comprising an inputting step of inputting the predetermined value.

35. A method according to claim 32, further comprising a priority inputting step of inputting priorities, wherein weighting on the similarities is performed in accordance with the priorities, in said determining step.

36. A computer-readable storage medium on which a program is encoded, the program comprising:

locus input code to input a locus;

recognition code to recognize an input image constituted by the locus in a plurality of recognition modes and to obtain first candidates of a recognition result in each of the plurality of recognition modes as a plurality of functions;

determination code to determine similarities of the first candidates so as to obtain second candidates of a recognition result, the second candidates having similar shapes to one another; and

selection code to select one of the second candidates from the plurality of functions as the recognition result.

37. A medium according to claim 36, wherein the input image is recognized using a plurality of recognition modes corresponding to types of an input image and wherein the second candidates are obtained by determining similarities in different modes of the plurality of recognition modes so that second candidates having different functions from one another are obtained.

38. A medium according to claim 37, wherein the plurality of recognition modes include a character recognition mode for recognizing an input image as a character and a command recognition mode for recognizing an input image as a command to edit characters of an input image.

39. A medium according to claim 38, wherein the program further comprises code to store similarities of the first candidates, wherein recognition is first performed in the command recognition mode, and if the recognition is successful, a corresponding similarity is referred to during selection of one of the second candidates.

40. A medium according to claim 37, wherein the program further comprises priority input code to input priorities of the respective recognition modes, wherein the second candidates are obtained by using the priorities.

41. A medium according to claim 37, wherein during execution of said recognition code, a predetermined order is added to the respective recognition modes, and if recognition in one mode is unsuccessful, recognition is performed in another recognition mode, in accordance with the order.

42. A medium according to claim 37, wherein the program further comprises storing code to store similarities of the first candidates, wherein recognition is first performed in the command recognition mode, and if the recognition is successful, a corresponding similarity is referred to in order to obtain the second candidates.

43. A medium according to claim 41, wherein if the recognition of input image is successful in one recognition mode, a recognized shape from the recognition is selected as a recognition result.

44. A medium according to claim 37, wherein the program comprises display code to display an image, wherein the first candidates which are recognized in the respective modes and have their similarities determined are displayed, for selection of one shape.

45. A medium according to claim 44, wherein the program further comprises reference input code to input a reference value of similarity, wherein the second candidates are displayed in accordance with the reference value.

46. A medium according to claim 36, wherein similarities between the input image and the first candidates are obtained and wherein a function based on one of the similarities is selected.

47. A medium according to claim 46, wherein if the difference between similarities of the first candidates is less than a predetermined value, the first candidates are displayed for selection of one shape.

48. A medium according to claim 47, wherein the program further comprises input code to input the predetermined value.

49. A medium according to claim 48, wherein the program further comprises priority input code to input priorities, wherein weighting is performed on the similarities in accordance with the priorities.

50. A computer-readable storage medium on which a program a is encoded, the program comprising:

locus input code to input a locus;

recognition code to recognize an input image constituted by the locus in a plurality of recognition modes and to obtain first candidates of a recognition result in each of the plurality of recognition modes;

judgment code to judge similarities in shapes among the first candidates obtained in the plurality of recognition modes;

display code to display, as second candidates, the shapes of the first candidates which are judged as having similar shapes; and

selection code to select one, as the recognition result, from the displayed second candidates, in accordance with a designation by a user.