Character recognizing method and apparatus

• US 5,812,696 A
• Filed: 06/04/1997
• Issued: 09/22/1998
• Est. Priority Date: 06/25/1992
• Status: Expired due to Term

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Harada, Takashi, Matsubayashi, Kazuhiro, Arai, Tsunekazu, Mori, Shigeki, Sakaguchi, Katsuhiko
• Primary Examiner(s): Kelley, Christopher S.
• Application Number:
  US08/869,223
• Time in Patent Office: 475 Days
• Field of Search: 382/177, 382/187, 382/192, 382/195, 382/197-203, 382/227, 382/185
• US Class Current: 382/185
• CPC Class Codes: G06V 30/242 Division of the character sequences into groups prior to recognition; Selection of dictionaries

Abstract

The invention has means for judging the characteristics of a character from a feature amount of one inputted character, thereby allowing a suitable character recognizer to be adapted to the input character in accordance with the feature amount of the input character. According to the invention, since the means for judging the characteristics of a character in accordance with the feature amount of the inputted character, a size of input character is not limited. Each suitable recognizer is naturally selected in accordance with the feature amount of the inputted character. The input character can be more accurately recognized.

Claims

1. A character recognition apparatus having a plurality of recognizing means, comprising:

input means for inputting a character pattern;

stroke extracting means for extracting strokes constructing a pattern from said inputted character pattern;

stroke curvature calculating means for calculating curvature of each of the strokes constructing the character pattern extracted by said stroke extracting means;

character curvature obtaining means for obtaining character curvature based on the stroke curvature for one character calculated by said stroke curvature calculating means; and

deciding means for deciding a proper one of said plurality of recognizing means to select on the basis of a comparison between the character curvature obtained by said character curvature obtaining means and a predetermined threshold value as a means for recognizing feature data extracted from the input character pattern.

2. An apparatus according to claim 1, wherein said input means inputs coordinate data of the character pattern.

3. An apparatus according to claim 1, further having character extracting means for extracting a pattern of one character.

4. An apparatus according to claim 1, further having coordinate input means.

5. An apparatus according to claim 1, further having pen input means for inputting coordinates.

6. An apparatus according to claim 1, further having display means for displaying the result of the recognition of said feature data extracted from the inputted character pattern.

10. An apparatus according to claim 1, wherein each of said plurality of recognizing means includes a dictionary which stores reference patterns.

12. An apparatus according to claim 1, wherein each of said plurality of recognizing means includes a dictionary which stores reference patterns.

7. A character recognizing method having a plurality of recognition algorithms, comprising the steps of:

inputting a character pattern;

extracting strokes constructing a pattern from said inputted character pattern;

calculating curvature of each of the strokes constructing the character pattern;

obtaining character curvature based on the stroke curvature for one character calculated in said calculating step; and

deciding a proper one of said plurality of recognizing algorithms to select on the basis of a comparison between the character curvature obtained in said obtaining step and a predetermined threshold value as a means for recognizing feature data extracted from the input character pattern.

8. A method according to claim 7, wherein said character pattern which is inputted is coordinate data.

9. A method according to claim 7, further having the step of extracting a pattern of one character from the inputted character pattern and, thereafter, extracting the strokes.

11. An apparatus according to claim 7, wherein each of said plurality of recognizing means includes a dictionary which stores reference patterns.

13. A computer-readable memory medium storing computer-executable process steps, the steps comprising:

an inputting step to input a character pattern;

an extracting step to extract strokes constructing a pattern from said inputted character pattern;

a calculating step to calculate curvature of each of the strokes constructing the character pattern;

an obtaining step for obtaining character curvature based on the stroke curvature for one character calculated by said calculating step; and

a deciding step to decide a proper one of said plurality of recognizing algorithms to select on the basis of a comparison between the character curvature obtained by said obtaining step and a predetermined threshold value as a means for recognizing feature data extracted from the input character pattern.

14. A computer-readable memory medium storing computer-executable process steps according to claim 13, wherein the character pattern which is inputted is coordinate data.

15. A computer-readable memory medium storing computer-executable process steps according to claim 13, the steps further comprising an extracting step to extract a pattern of one character from the inputted character pattern and an extracting step to extract the strokes.