| US 7,469,063 B2 | ||
| Apparatus, method and storage medium storing program for recognizing characters | ||
| Toshiya Koyama, Nakai-machi (Japan); Teruka Saito, Nakai-machi (Japan); Masakazu Tateno, Nakai-machi (Japan); Kei Tanaka, Nakai-machi (Japan); Masayoshi Sakakibara, Ebina (Japan); and Kotaro Nakamura, Minato-ku (Japan) | ||
| Assigned to Fuji Xerox Co., Ltd., Tokyo (Japan) | ||
| Filed on Mar. 15, 2005, as Appl. No. 11/79,178. | ||
| Prior Publication US 2006/0078200 A1, Apr. 13, 2006 | ||
| Int. Cl. G06K 9/00 (2006.01); G06K 7/10 (2006.01); G06K 9/20 (2006.01); H04N 1/04 (2006.01) | ||
| U.S. Cl. 382—187 [382/321; 358/474; 358/498] | 12 Claims |
| 12. An apparatus for recognizing characters, comprising:
a text structure analyzing unit that analyzes an input image of text comprising mixed typed and handwritten characters and
separates the input text image into a plurality of text imaging regions;
a feature vectors calculating unit that calculates feature vectors of the text imaging regions in terms of a centroid or a
center of a circumscribed quadrangle of a character, respectively;
a feature vectors aggregating unit that aggregates the feature vectors for each of the text imaging regions over the entire
input text image, and calculates a separation coefficient for separating typed characters and handwritten characters, based
on the result of the aggregation of the feature vectors; and
a typed and handwritten character separation unit that separates characters involved in the input text image into the typed
characters and the handwritten characters, based on the separation coefficient.
|