JPS63279386A - Character recognizing device - Google Patents

Abstract

PURPOSE:To easily discriminate the characters of the same shape, which are difficult to be distinguished only from the font of display characters by voice- outputting reading information corresponding to recognition candidate characters and correcting the recognition result of the characters of the same shape by using the read information. CONSTITUTION:A storage means 8 in which reading information of the characters for recognition are stored is provided. When the recognition result is different from the correct character, or it is difficult to discriminate whether the character is correct or not due to the characters of the same shape, an operator uses an input means 6 and instructs the display of the recognition candidate characters for a display means 5 to control means. One character among the recognition candidate characters displayed in the display means 5 is designated by a carsor or the like. At that time, the control means 7 reads reading information corresponding to the designated character, transmits it to a voice output means 9 which outputs the reading of the designated character as voice. The operator collates information of the characters outputted as voice and selects and inputs the correct character among the recognition candidate characters displayed in the display means 5.

Description

[Detailed Description of the Invention] [Industrial Application Field] This invention relates to a character recognition device that recognizes many types of characters, and particularly to a character recognition device that can easily correct unrecognized characters and misread characters. It is.

[Prior Art] FIG. 2 is a block diagram of a conventional character recognition device shown in, for example, the "Mitsubishi OCRM2483-N Overview." In the figure, (1) is a form, (2) is a scanning means, ( 3) is a recognition means, (4) is a recognition dictionary storing reference patterns of characters to be recognized, (5) is a display means, (6) is an input means, (7)
is a control means.

FIG. 3 is an explanatory diagram of a screen display when correcting the recognition result of conventionally used character recognition.

In FIG. 3, (9) is a recognition result, (10) is a pattern corresponding to a character to be corrected, (11) to (14) are recognition candidate characters for the character to be corrected, and (15) is a cursor.

Here, the symbol "◆" as the recognition result (9) is used as a symbol indicating an unrecognizable character.

Next, the operation will be explained. Characters on a form (1) are photoelectrically converted by a scanning means (2), and a human character pattern is sent from the scanning means (2) to a recognition means (3). Recognition means (3)
Then, the input character pattern is compared with the reference pattern in the recognition dictionary (4) to recognize the character, and the recognition result is sent to the control means (7).
).

If the recognition result is different from the correct character, the recognition candidate character is displayed on the display means (5), and the operator selects and inputs the correct character using the input means (6) such as a keyboard.

However, when the types of characters to be recognized include kanji, hiragana, katakana, symbols, etc., multiple characters with exactly the same shape (hereinafter referred to as homomorphic characters) may be included.

Words including the preceding and following characters, without using grammatical information,
Correct recognition from only a single character pattern information requires a high degree of skill, although it is not impossible. Therefore, even in a character recognition device, if such isomorphic characters are recognized, there is a high probability that the characters will be unrecognized or misread, and correction input will be performed using recognition candidate characters.

Examples of isomorphic characters are ``力'' (Kanji) and ``力'' (Katakana), ``工'' (Kanji) and ``工'' (Katakana), and ``-
” (kanji ichi), “−” (long sound symbol), and “−”
(negative sign) and “-” (hyphen), “he” (hiragana)
and “he” (katakana).

In Figure 3, the recognition result (9) for “-” (long sound symbol) is recognized (◆ in the figure), and the character pattern (10) and recognition candidate characters “-” (Kanji one) (11), “ −”
(minus sign) (12), “-” (major sign) (13),
“-” (hyphen) (14) is displayed.

When modifying a character, the operator moves the cursor (15) to the correct character position and inputs this character. In this case, although there are subtle differences in the display fonts of the recognition results and recognition candidate characters, it requires a high degree of skill, if not impossible, for the operator to identify all of these and select and input them correctly. shall be.

[Problems to be Solved by the Invention] Since the conventional character recognition device is configured as described above, an operator must identify isomorphic characters based on differences in display fonts, and in order to perform the identification process, This has the problem of causing extreme fatigue and making it easy to make incorrect inputs.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a character recognition device that can improve the operability of correction work and significantly reduce operator fatigue.

[Means for Solving the Problems] ゛The character recognition device according to the present invention is provided with a storage means that stores information on the pronunciation of the character to be recognized, and reads the pronunciation of the character designated with a cursor or the like among the recognition candidate characters by voice. The output means outputs audio.

[Operation] The character recognition device according to the present invention outputs the pronunciation of the character specified by the cursor or the like among the recognition candidate characters, thereby identifying isomorphic characters that are difficult to distinguish only from the font of the character displayed on the display means. Easy to identify, reduces operator fatigue and reduces input errors.

[Embodiment] An embodiment of the present invention will be described below with reference to FIG. 1, in which the same parts as in FIG. 2 are given the same reference numerals.

In FIG. 1, (8) is a storage means that stores reading information corresponding to each of the character types to be recognized, and (9) is an audio output that outputs audio according to the reading information stored in the storage means (8). It is a means.

Next, the operation will be explained using FIGS. 1 and 3. The operation up to displaying the recognition result on the display means (5) is the same as that of the conventional apparatus shown in FIG. 2, and therefore the explanation will be omitted.

Here, if the recognition result is different from the correct character, or if it is an isomorphic character and it is difficult to determine whether it is the correct character or not,
The operator uses the input means (6) to display the display means (5).
The control means (7) is instructed to display the recognition candidate characters on the screen.

Further, the operator specifies one of the recognition candidate characters displayed on the display means (5) using a cursor or the like.

At this time, the control means (7) reads reading information corresponding to the specified character from the storage means (8) and sends it to the voice output means (9). Then, the audio output means (
9), for example r PARCOR type speech analysis and synthesis system J
(NTT Tsuken Research Practical Application Report, 27.6.1061
-1078.1978) and other well-known PAs reported in
Using the RCOR method, etc., the reading of the specified characters is output as audio.

The operator selects and inputs the correct character from among the recognition candidate characters displayed on the display means (5) by referring to the information on the reading of the character output as voice.

For example, in Figure 3, the recognition result for “-” (articulation symbol) (
9) becomes unrecognizable (◆ in the figure), and the cursor (15)
is specified in the place of the recognition candidate character "-" (long sound) (15), so the voice output means (9) outputs the voice "Cho-on". Therefore, the operator can determine that the character corresponding to the position of the cursor (15) is a long sound symbol, and can easily select and input even if there is an isomorphic character among the recognition candidate characters.

In addition, in the above embodiment, the case where the pronunciations are output aloud for all the recognition candidate characters specified with the cursor etc. has been explained, but only when a recognition candidate character with isomorphic characters is specified, the reading of the corresponding character is output. The reading may be outputted aloud, and in this case, the memory size of the storage means can be significantly reduced.

[Effects of the Invention] As described above, according to the present invention, the reading information corresponding to the recognition candidate character is outputted by voice, and the recognition result of the isomorphic character is corrected using this reading information. Therefore, it is possible to easily identify isomorphic characters that are difficult to distinguish only from the font of displayed characters, which has the effect of reducing operator fatigue and reducing input errors.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a character recognition device according to an embodiment of the present invention, FIG. 2 is a block diagram showing a conventional character recognition device, and FIG. 3 is an explanatory diagram of a screen display when a recognition result is corrected. In the figure, (1) is a form, (2) is a scanning means, and (3)
is a recognition means, (4) is a recognition dictionary, (5) is a display means, (
6) is an input means, (7) is a control means, (8) is a storage means, and (9) is an audio output means. In addition, in the figures, the same reference numerals indicate the same or equivalent parts. Agent Patent attorney Masuo Oiwa (and 2 others) Figure 1 Figure 2 Figure 3-9: Recognition result 1o: Pattern corresponding to correction target character 11: Recognition candidate character (Kanji no Ichi) 12: Recognition candidate character (Negative sign) 13: Recognition candidate character (long sound symbol) 14: Recognition candidate character (-j' 7n) Procedural amendment 3-2) Showa year month/day 62'+215 Dear Commissioner of the Japan Patent Office
x7. ：、：；、31、Indication of the incident Tokkō Sho 6
2-115105 No. 2, Character recognition of the name of the invention, recognition device 3, relationship with the case of the person making the amendment Patent applicant representative Moriya Shiki 4, agent address 2-2-3 Marunouchi, Chiyoda-ku, Tokyo ~
447″ 5. Detailed description of the invention in the specification subject to amendment. 6. Contents of l+l1il

Claims (1)

[Claims] A scanning means for scanning and photoelectrically converting characters recorded on a form, a recognition dictionary storing a reference pattern of characters to be recognized, and character recognition by comparing an input character pattern from the scanning means with the reference pattern. a recognition means for displaying the recognition result of the recognition means and a recognition candidate character, a storage means for storing information on the reading of the recognition target character, and a reading of the recognition candidate character displayed on the display means. A character recognition device comprising: a voice output means for outputting a voice according to reading information stored in the storage means; and an input means for selectively inputting a character from among the recognition candidate characters displayed on the display means.