JPH0757041A - Character recognition device - Google Patents

Abstract

PURPOSE:To provide the character recognition device which can easily find a misrecognized character. CONSTITUTION:The character recognition device 1 consists of a read part 2 which reads out characters entered in a body to be read and recognizes the characters, a control part 3 which outputs the recognition result obtained by the read part 2 and outputs a corrected character than an operator inputs as the recognition result in the case of misrecognition, and a result output part 4 which informs the operator of the recognition result sent from the control part 3; and the control part 3 is equipped with a sound signal generation part 34 which generates a sound signal including the characters of the recognition result consisting of recognized characters and corrected characters and the result output part 4 is equipped with a sound output part 43 which outputs the sound signal sent from the sound signal generation part 34 in the form of a voice.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition device for optically reading and recognizing a character written on an object to be read, such as a form or a general document.

[0002]

2. Description of the Related Art A character recognition device composed of an OCR or the like is for optically reading and recognizing a character written on an object to be read such as a form or a general document. This is intended to improve the efficiency of. FIG. 3 is a configuration diagram illustrating a conventional character recognition device. That is, the character recognition device 1 mainly includes a reading unit 2 for reading characters written on an object to be read, a control unit 3 for receiving a character code obtained as a recognition result and outputting the character code, and a control unit. 3 and a result output unit 4 for notifying the operator of the recognized character received from the character code output from the character output unit 3.

In order to recognize a character written on an object to be read using this character recognition device 1, first, the reading section 2 is used.
An object to be read, for example, a form is input to the input unit 21 of the above, and the reflected light based on the characters written on the form is converted into an electric signal by the photoelectric conversion unit 22 including a CCD sensor or the like. Next, the recognition unit 23 cuts out an image of a certain area formed by the electric signal, extracts the characteristics of characters in the image, and collates the characteristics with a dictionary stored in advance to recognize the characters.

The recognized result is a character code, and the control unit 3
And the recognition result sending unit 33 sends the character code to the result output unit 4. The result output unit 4 receives the character code and outputs the character code as a character by the display unit 42 such as a display. In addition, the character recognized by the character output unit 41 is output to paper or the like as needed.

If, for example, an erroneous recognition is made in the characters displayed on the display unit 42, the operator finds it and inputs the correct character from the keyboard unit 35 to rewrite the display on the display unit 42. . For example, JP-A-5
In the optical character reader disclosed in Japanese Patent Laid-Open No. 9-36878, the image signal of the read character and the discrimination result of the character are combined and displayed on a display or the like so that the operator can easily find the portion erroneously recognized by the apparatus. Has become.

[0006]

However, such a character recognition device has the following problems. That is, in this character recognition device, since the images of the recognized characters are displayed on the screen of the display or the like in the order corresponding to the scanning order, it is necessary for the operator to visually find the erroneously recognized character that needs correction from this screen. There is.

In the optical character reader disclosed in Japanese Patent Laid-Open No. 59-36878, the image signal of the read character is combined with the character discrimination result and displayed on the display or the like. Although it is easy to find out whether or not it can be identified as a character, for example,
With regard to characters having the same shape but different character types, such as katakana “ka” and kanji “power” (power), the operator has a high possibility of overlooking even if the recognition is incorrect because the display is similar.

Therefore, it is necessary for the operator to visually check the display on the display to confirm whether or not the character can be discriminated and also to confirm the difference in the character type. Will be required. Therefore, an object of the present invention is to provide a character recognition device that can easily detect a misrecognized character.

[0009]

SUMMARY OF THE INVENTION The present invention is a character recognition device made to solve such problems. In other words, this character recognition device outputs a recognition result obtained by the reading unit and a reading unit for optically reading a character written on an object to be read and performing character recognition based on the read image. A control unit for outputting the correction character input by the operator at the time of recognition as a recognition result, and a result output unit for notifying the operator of the recognition result sent from the control unit,
The control section is provided with a voice signal generating means for generating a voice signal corresponding to the character in the recognition result, and the result output section is provided with a voice output section for outputting the voice signal sent from the voice signal generating means as a voice. It is a thing.

Further, it is also a character recognition device for generating a voice signal containing a correction character input by an operator by a voice signal generating means. Further, when the operator inputs a correction character, the voice signal generating means displays the beginning of the correction character, the beginning of the sentence including the correction character, the beginning of the word including the correction character, the beginning of the phrase including the correction character, and the correction character. It is also a character recognition device that regenerates a voice signal from any of the beginnings of paragraphs that include it.

[0011]

The voice signal generating means provided in the control section sends a voice signal corresponding to the recognized character to the result output section, and this voice signal is output as voice by the voice output section provided in the result output section. Then, the recognized character can be notified to the operator as a voice. In addition, a voice signal containing the correction character input by the operator, that is, a voice signal is generated again from the voice signal generating means from the beginning of any of the beginning of the correction character and the sentence, word, clause, or paragraph including the correction character, By outputting the corrected character as voice by the output unit, the operator can confirm the corrected character by voice.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the character recognition device of the present invention will be described below with reference to the drawings. FIG. 1 is a configuration diagram illustrating a character recognition device of the present invention. The character recognition device 1 mainly includes a reading unit 2 for reading characters written on an object to be read, a control unit 3 for obtaining a recognition result from the reading unit 2 and outputting a character code, and a character code for the control unit 3. Result output unit 4 which informs the operator of the recognized characters
And a voice signal generating unit 3 in the control unit 3.
4 is provided, and the result output unit 4 is provided with the voice output unit 43.

It should be noted that the same reference numerals as those in FIG. 3 among the reference numerals shown in FIG. 1 denote the same components, and in the following, the description will be focused on the new portions. That is,
The control unit 3 of the character recognition device 1 includes a readable / writable memory 31 for temporarily storing the recognition result of the character sent from the recognition unit 23 of the reading unit 2, and via the memory 31. The recognition results are sent to the recognition result sending section 33 and the voice signal generating section 34, respectively.

Further, the correction unit 32 of the control unit 3 includes a memory 3
A character that needs to be corrected is detected from the character string data stored in No. 1 and the character is corrected and the corrected character string data is sent to the voice signal generator 34.

The voice signal generator 34 receives the character string data stored in the memory 31, generates a voice signal corresponding to the character, and sends the voice signal to the result output unit 4. The voice output unit 43 provided in the result output unit 4 is composed of, for example, a speaker, and converts the voice signal sent from the voice signal generation unit 34 into voice to notify the operator of the recognized character as voice. . Therefore, the operator can check the recognized character by looking at the display unit 42 including the display of the result output unit 4 and at the same time, hear the character as a voice.

For example, when the character cannot be read because the reading is incomplete, the character is not displayed on the display, or the character cannot be read, so that the operator finds the character and judges the recognition error. Also, for example, katakana ka and kanji kanji (power)
Characters having the same shape but different character types are difficult to determine only by the display on the display, so the determination is made by the voice from the voice output unit 43 generated simultaneously with the display.

As described above, the operator is the character recognition device 1
Confirm whether or not the recognized character is erroneously recognized by the display unit 42 and voice from the voice output unit 43, and if there is erroneous recognition, correct it by inputting the correct character from the keyboard unit 35. To do. The corrected character is displayed on the display unit 42, and the voice signal including the corrected character is output from the voice output unit 43 as voice again.

Next, a series of operations of the character recognition device 1 will be described with reference to the flowchart of FIG. Note that FIG.
Reference numerals not shown in FIG. 1 refer to FIG. First, the recognition result of the reading unit 2 is stored in the memory 31, and the character string data of the recognition result is sent to the recognition result sending unit 33 and the voice signal generating unit 34.

Then, the character string data is sent from the recognition result sending section 33 to the display section 42 to display the recognized characters on a display or the like. At the same time as the display, a voice signal corresponding to the character string data is generated from the voice signal generation unit 34 and sent to the voice output unit 43 to output the character as voice from a speaker or the like.

The operator confirms the recognized character based on the display and the voice from the speaker, and if there is an erroneous recognition, sends a signal to the control unit 3 to stop the voice output by operating the keyboard unit 35. The control unit 3
When this signal is received, the determination in step 2a is “Y
"es", the voice output is interrupted in the next step 2b.

After that, in the next step 2c, the operator uses the keyboard section 35 to correct the erroneously recognized character. With this correction, the correction unit 32 rewrites the erroneous recognition portion of the character string data stored in the memory 31, and also sends the character corrected in step 2d to the voice signal generation unit 34.

Next, in step 2e, the voice signal generator 34 extracts the character string data after the corrected character from the memory 31 starting from the sent corrected character. As the extraction of the character string data, the punctuation mark immediately before the correction character is detected to extract the character string data including the correction character with the character next to the punctuation point as a starting point, or the sentence including the correction character, You may make it extract any one of a word, a clause, and a paragraph.

Then, in the next step 2f, the previously extracted character string data is converted into a corresponding voice signal, and the voice signal is sent to the voice output section 43. And step 2g
As a restart of the voice output, the voice output unit 43 outputs voice based on the voice signal, displays the corrected character on the display, and outputs the voice signal including the corrected character again to notify the operator.

As a result, the operator confirms the corrected character, and after confirming it, similarly confirms the next recognized character string. By repeating this series of steps, it is possible to accurately confirm the recognized character and correct the erroneous recognition.

[0025]

As described above, the character recognition device of the present invention has the following effects. That is, since the recognition result is displayed on a display unit such as a display and is output as a voice from a voice output unit such as a speaker, a misrecognized character can be easily found. In particular, when confirming characters of the same shape but different character types, it is possible to surely find them by listening to the sound even if they cannot be easily found only by the display. Also, when the operator inputs a correction character, the character string immediately before including the correction character is output by voice again, so that the operator can confirm the correction character without reading back the character string by itself. It will be possible to make good corrections.

[Brief description of drawings]

FIG. 1 is a configuration diagram illustrating a character recognition device of the present invention.

FIG. 2 is a flowchart illustrating a series of operations of the character recognition device of the present invention.

FIG. 3 is a configuration diagram illustrating a conventional character recognition device.

[Explanation of symbols]

 1 Character Recognition Device 2 Reading Unit 3 Control Unit 4 Result Output Unit 34 Voice Signal Generating Unit 35 Keyboard Unit 42 Display Unit 43 Voice Output Unit

Claims (3)

[Claims] 1. A reading unit for optically reading a character written on an object to be read and recognizing the character based on the read image, and a recognition result obtained by the reading unit and for erroneous recognition. A character recognizing device comprising a control unit for outputting a correction character input by the operator as the recognition result at the time, and a result output unit for notifying the operator of the recognition result sent from the control unit. The unit includes a voice signal generating unit that generates a voice signal corresponding to a character in the recognition result, and the result output unit is a voice output unit for outputting the voice signal sent from the voice signal generating unit as a voice. A character recognition device characterized by being provided. 2. The character recognition device according to claim 1, wherein the voice signal generating means generates a voice signal containing a correction character input by the operator. 3. The character recognition device according to claim 2, wherein when the operator inputs a correction character, the voice signal generating means includes the correction character, a sentence including the correction character, a word including the correction character, A character recognition device characterized in that the voice signal is regenerated from the beginning of either a clause including a correction character or a paragraph including the correction character.