JPS594358Y2 - Character control device in character correction - Google Patents

Description

[Detailed description of the invention] When an optical character reading device reads out a difficult-to-read character or a character that may be misread, the present invention enlarges or reduces the character or character string containing the character and displays it for decipherment. The present invention relates to a character control device for character correction.

Conventionally, in an optical character reading device (hereinafter referred to as OCR), a method is used in which characters on a form, including handwritten characters, are compared with a dictionary or the like for character recognition.

Here, when a difficult-to-read character or a character that may be misread due to post-processing checks is read, this character is
There is a method in which the data is displayed on the display of the CR, read by the operator, and inputted using keys from the keyboard to make corrections.

In this case, the size of the character image displayed on the display device was determined by the internal specifications of the device, and the operator could not control the size of the image, which sometimes made it difficult to read.

However, in the case of handwritten characters, the size varies depending on the characteristics of the person writing it, so when an operator reads the image on the display device, if it is small and difficult to read, it can be enlarged, and if it is difficult to read with just a single character, it can be enlarged. In some cases, it is desirable to be able to display images of the preceding and following characters and to be able to reduce the size in order to make judgments based on the context of the characters.

An object of the present invention is to provide a character control device for character correction that makes it easier for an operator to read difficult-to-read characters in character recognition.

In order to achieve the above object, the character control device for character correction of the present invention scans the characters written on the form with an optical character reading device, and uses the image information of the characters to identify difficult-to-read characters or characters by character recognition. When a character is read as a character that may be misread as a result of the post-processing check, at least the video information of the character corresponding to the character is extracted and displayed on a display device, and an operator reads it and inputs data from a keyboard. The character correction device enlarges the image information of the extracted character at a ratio determined by instructions given by an operator, or enlarges the image information of a plurality of character strings before and after the character including the extracted character. The present invention is characterized in that it includes means for displaying reduced information.

The present invention will be described in detail below with reference to examples.

FIG. 1 is a schematic explanatory diagram of an optical character reading device to which the present invention is applied.

In the figure, a form to be subjected to character recognition is sent to a readable position by the operation of the paper feed mechanism 4 and the control unit 3,
Through the operation of the scanning mechanism 6 and the control unit 5, character video information is written into the video storage unit 7, and the recognition unit 8 analyzes the information by extracting the characteristics of this information and comparing it with a dictionary.
The recognition result is stored in the data storage area of the main storage unit 2.

These series of operations are performed by executing a program stored in the main storage section 2 under the control of the CPU 1.

Furthermore, if the program inspects the recognition results and determines that, for example, there are difficult-to-read characters and correction by image display is necessary, the program stores video information of the corresponding characters in the memory provided in the control unit 9 of the display device 10. and display it.

In the present invention, video information is compressed according to the size of characters displayed on the screen.

This compression ratio is divided into several stages, and the keyboard 1
2 and the same control unit 11 corresponds to the operator's keystroke information.

That is, information such as whether the type of character to be read is handwritten or printed is stored in the main memory 2 in advance, and the size of the image based on that information is set as a reference value at a certain stage of the compression ratio described above. The size of the image is determined and displayed, and the size of the image is made variable by key operations by the operator.

FIGS. 2a to 2f are functional explanatory diagrams of the main parts of the present invention.

In the figure a, the paper feed direction 15 of the form 20, the scanning direction 1
Set it to 6.

Now, assume that handwritten characters 13 and printed characters 14 are written on the form 20, and in each of these numbers "123...", "3" is a difficult to read character or a character that may be misread. .

As shown in the figure, the difficult-to-read character "3" among the handwritten characters 13 is displayed in the standard size determined by the device as described above, and when the operator wants to enlarge it and read it,
The operator presses the screen control key on the keyboard to enlarge the display.

If it is desired to show the context of the difficult-to-read character "3", the characters before and after are displayed and the corresponding character is indicated with the cursor 17, as shown in FIG.

Figure d displays a plurality of characters, including characters that may be misread in the printed characters 14, and Figure f shows a portion of them enlarged and the corresponding character is indicated with the cursor 17.

FIG. 3 is an explanatory diagram of an example of a specific circuit particularly related to the functions of the present invention in the optical character reading device of FIG. 1.

Correspondence with FIG. 1 is indicated by a broken line. In the figure, a reading control table 21 in the main storage section 2 stores character types, number of characters, address information, etc. related to character reading and image display, and supplies them to each related circuit.

By this control, the character scanning circuit 2 in the scanning mechanism control section 5
The character video information output from 2 is multi-level converted into a digital signal via the image information A/D conversion circuit 23, and further stored in the video storage section 7 consisting of the image memory 25 via the binarization conversion circuit 24. do.

Next, in the recognition section 8, the image information stored in the image memory 25 is input into the recognition circuit 26, and character recognition is performed by positioning, feature extraction, comparison with a dictionary, etc., and the recognition result is stored in the recognition result storage buffer 27. The post-processing circuit 28 checks the content for difficult-to-read characters or characters that may be misread, and sends the video information of the checked-out characters to the image display processing circuit 29 of the display device control section 9, where it is processed using the method shown in FIG. As explained above, when difficult-to-read characters are small and difficult to judge, they are enlarged, and when it is necessary to judge from the context, multiple character strings containing difficult-to-read characters are reduced and displayed, and the operator deciphers this.

As a result, the operator inputs the corrected character information into the keyboard input information register 30 in the keyboard control unit 11 by pressing keys on the keyboard, and the contents are transferred to the image display processing circuit 29.
to correct the aforementioned checkout characters.

This result is stored in the output data buffer 32 via the editing and output processing circuit 31.

Character correction is performed in this manner.

FIG. 4 shows an image display processing circuit 29 which is the main circuit of the present invention.
A detailed circuit example is shown below.

In the figure, right side circuits 41 to 47 indicate each control circuit that follows the processing procedure of the image display processing circuit 29, and left side circuits are other circuits shown in FIG. 3, and indicate registers or buffers that exchange signals with the right side circuit. .

That is, the checkout character information output from the post-processing circuit 28 in FIG.
The checkout character address calculation circuit 42 specifies the address of the checkout character using the address information from the reading control table 21, and performs image display compression. The compression ratio is input to the ratio determination circuit 43 and determined by information set by the operator from the keyboard input information register 30, and the image display circuit 45 is inputted to the image display conversion processing circuit 44 which inputs image information from the image memory 25. and send the expanded or compressed character information to screen buffer 4.
8 and displayed in the same manner as described above.

Based on this display, the operator deciphers the enlarged or compressed checkout character, inputs the operator's correction information from the keyboard input information register 30 via the correction key manual processing circuit 46, corrects it in the recognition result correction processing circuit 47, and recognizes it. The result is stored in the result storage buffer 27.

Note that various known methods can be applied to enlarge or compress image information.

As explained above, according to the present invention, when a difficult to read character or a character that may be misread is read by OCR, the character is enlarged or the character string containing the character is reduced and displayed. It is designed to be readable.

By doing so, it becomes easier for the operator to read the characters, and the characters can be easily and accurately corrected.

[Brief explanation of the drawing]

Figure 1 shows an optical character reader (OCR) to which the present invention is applied.
2 a-f are functional explanatory diagrams of the main parts of the present invention, Fig. 3 is an explanatory diagram of a specific circuit example related to the functions of the present invention, and Fig. 4 is the main part of Fig. 3. It is a detailed explanatory diagram of the circuit, in which 1 is a CPU, 2 is a main memory section, 5 is a scanning mechanism control section,
1 is a video storage unit, 8 is a recognition unit, 9 is a display control unit, 10 is a display device, 11 is a keyboard control unit, 12 is a keyboard, 13 is a handwritten character, 14 is a printed character, 17 is a cursor,
20 is a form, 21 is a reading control table, 22 is a character scanning circuit, 23 is an image information A/D conversion circuit, 24 is a binarization conversion circuit, 25 is an image memory, 26 is a recognition circuit, and 27 is a recognition result storage buffer. , 28 is a post-processing circuit, 29 is an image display processing circuit, 30 is a keyboard input information register, 31
3 is an editing and output processing circuit, 32 is an output data buffer, and 48 is a screen buffer.

Claims (1)

[Scope of utility model registration request] When the image information of the characters obtained by scanning the characters written on the form with an optical character reader is read as difficult to read characters by character recognition or characters that may be misread by post-processing check, A character correction device that extracts at least video information of the character corresponding to the character, displays it on a display device, and performs correction by an operator reading it and inputting data from a keyboard, according to instructions given by the operator. A character correction characterized by comprising means for enlarging the image information of the extracted character at a predetermined ratio or reducing and displaying the image information of a plurality of character strings before and after the character including the extracted character. Character control device in.