JPH0762859B2 - Character pattern recognition and correction device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character reading device for Kanji OCR and the like, and in particular, character pattern recognition that can be easily corrected when misreading or misreading occurs during recognition. Revision device.

[Background of the Invention]

OCR is a method of scanning the characters printed or handwritten on a form by means such as photoelectric conversion, analyzing the obtained image data, and recognizing it to determine what kind of character it is, and the corresponding character code. Is a device for generating. However, in this type of device, unreadable characters or erroneous reading occurs due to stains on the form, print quality, or disorder of the handwritten typeface. In such a case, the operator usually inputs the correct character code by key operation while referring to the display screen and the form, and replaces the previously generated character code with this. However, this correction method has a drawback in that it is difficult to make correspondence between the characters on the display screen and the characters on the form, and the work efficiency is poor.

Therefore, in the conventional device, as described in Japanese Patent Laid-Open No. 59-160273, the character data corresponding to the character code of the read result and the grayscale image data of the form are simultaneously displayed on the display. In addition, many conventional products display binary image data at the same time instead of the grayscale image.

According to this method, the character pattern information for a plurality of lines can be displayed, and the correction work can be effectively performed. However, no consideration was given to the following two points.
The first point is that with the conversion of OCR to kanji, the size of the entered characters tends to increase, which in turn tends to increase the size of the form.
Therefore, even if the resolution of the image data is reduced, it is difficult to display one original image on the display. The second point is that with the cost reduction of the device, unlike the conventional OCR with a display correction function, the OCR shares only the recognition function, and the display correction of the recognition result mainly uses a high-end machine such as a personal computer. Tend to occupy. On the other hand, the communication function of the host machine for receiving data from the OCR is often that of a low speed serial interface. Therefore, unlike code data, it becomes difficult to transfer image data having a large amount of data at a high speed in accordance with the correction speed, and it is impossible to improve the correction efficiency.

[Object of the Invention]

The object of the present invention is to solve the above-mentioned problems of the prior art by displaying the recognition result of the character pattern drawn on the document and the layout information indicating the mutual relation between the character pattern such as the character frame and the ruled line on the document. It is an object of the present invention to provide a character pattern recognition / correction device capable of easily correcting at least a part of a character pattern by displaying it in another area above and referring to the layout information display.

[Outline of Invention]

In order to achieve the above object, the present invention provides a character reading unit that captures at least a character pattern from a document having a character pattern and layout information surrounding the character pattern to obtain an image signal of the character pattern, and a character reading unit on the document. Storage means for storing layout information, and an image signal of the character pattern imaged by the character reading means is cut out based on format information obtained from the layout information stored in the storage means, and the character pattern is recognized and stored. Recognizing means, display means for displaying the layout information and the character pattern recognized and stored by the recognizing means in different areas on the screen so that the layout information stored in the storing means can be referred to; And a correction means for correcting at least a part of the character pattern displayed on the screen of the display means. A character pattern recognition correcting device according to symptoms. According to the present invention, in the character pattern recognition / correction device, the display means is configured to highlight a range in the display area of the layout information corresponding to a range of the character pattern in the display area of the character pattern. Is characterized by.

Therefore, the recognition result of the character pattern drawn on the document and the layout information indicating the mutual relationship between the character patterns such as character frames and ruled lines on the document are displayed in another area on the screen and the layout information display is referred to. However, at least a part of the character pattern can be easily modified.

Example of Invention

First, the principle of character pattern recognition and correction according to the present invention will be described with reference to FIGS. 1 (a) and 1 (b). The present invention reads the image signal of the input form 1 (Fig. 1 (a)) which is the target of character reading, recognizes the pattern, obtains the code information thereof, and the format information stored in the personal computer for result correction. The layout screen of the form is obtained from the display 2 of the personal computer (FIG. 1 (b)), and the layout screen 21 corresponding to the character pattern 22 corresponding to the character code of the read result and the two-dimensional image of the input form is displayed. The format information is used for displaying the layout information and the information that specifies the position of the character frame and the type of characters to be input because the character pattern on the document (form) is input as the format information. This refers to position information such as ruled lines, and this format information is usually personal computer due to memory capacity limitation of OCR. It is stored on the host machine side such as a computer and is transferred to the OCR side when necessary.The OCR cuts out the character pattern according to the former information and recognizes the character pattern by switching the target character type and recognizing the character pattern. On the other hand, the layout information display is composed of simple symbols and figures, that is, in the present invention, the character data such as the read result is displayed as usual, but the outline of the form is displayed. Is displayed on the form layout screen from the format information, so that the operator can immediately determine which character on the form should be corrected by looking at the layout screen, and the correction data can be easily obtained. It becomes possible to enter a key.

Further, according to the present invention, the following method becomes more effective. That is, a function is added to display only the partial area 23 corresponding to the character pattern 22 of the read result displayed on the display 2 on the layout screen 21 by changing the color tone or the brightness (hatched portion in FIG. 1B). To do. That is, the reading result to be corrected displayed on the display is
The position corresponding to the form is displayed by changing the color tone and brightness. Therefore, the operator can more easily key in the correction data by referring to the corresponding area of the input form.

Furthermore, you can reflect the cursor position of the character pattern corresponding to the character code of the read result on the layout screen,
Needless to say, the correction can be further facilitated by displaying the candidate character group 24 or the like corresponding to the cursor position. Here, the candidate character group refers to a character group having a high certainty as a result of character reading. For example, when a gate is input, a similar interview or the like is output as a candidate character group.

Next, specific examples of the present invention will be described with reference to the drawings.

FIG. 2 is a block diagram showing the configuration of an embodiment according to the present invention.

In FIG. 2, the character group entered on the form 3 is photoelectrically scanned by the raster scanning unit in the character observing unit 4 of the OCR 41 and binarized. Further, in the character observing section 4, one character of the obtained image signal is cut out according to the format information stored in the format storing section 7,
It is sent to the recognition unit 5. Character recognition is performed in the recognition unit 5, and the character code is stored in the result buffer 6. The character code stored in the result buffer 6 is transferred to the result buffer 9 created in the storage area of the personal computer 91 in response to the request of the upper personal computer 91. On the other hand, the format information used for cutting out the characters is the personal computer of the higher rank.
In addition to being created by 91, it is usually stored in the format storage section 13 of the personal computer 91 due to the memory capacity limitation of the OCR 41, and is transferred to the format storage section 7 of the OCR 41 as necessary. Therefore, input form 3
The format information is stored in the format storage section 13 of the upper personal computer 91 used for display correction. The format information of the input form 3 stored in the format storage unit 13 is sent to the layout creating unit 14. The layout creating unit 14 creates the layout screen 21 shown in FIG. 1B based on the format information.
This image signal and the output signal of the character code that has been formatted in the result buffer are sent to the screen creating section 10. Both are combined in the screen creating unit 10, converted into a video signal, and then sent to the display 11. On this display 11, as shown in FIG. 1 (b), the character of the recognition result is output on the right side of the screen, and the layout for one form is displayed on the left side of the screen. The operator refers to the input form and the layout screen and operates the keyboard 12 to correct the display data on the right side. In this case, the layout screen and the display area of the recognition result may be reversed left and right. It is also possible to display or delete the layout screen at an arbitrary place as needed.

FIG. 3 shows a recognition controller 8 and a display correction controller 15
3 is a flow chart showing an operation example of the above. First, the recognition controller 8 drives the character observing unit 4 to scan and binarize a character group for one form (31). One character of this binarized signal is sent to the recognition unit 5 which is cut out according to the format information in the format storage unit 7. Next, when the recognition controller 8 drives the recognition unit 5, the recognition unit 5 sequentially executes the recognition process character by character and stores the recognition result in the result buffer 6 (32). In this way, when the character recognition for one form or one line is completed, the recognition controller 8 informs the display correction controller 15 to that effect. When the display correction controller 15 receives the end signal, the result buffer 6
The contents are transferred to the result buffer 9 and the layout creating section 14 is driven to create a layout screen from the format information of the input form stored in the format storage section 13. Here, the layout screen is
It can be created in advance. Next, when the display correction controller 15 drives the screen creation unit 10, the result buffer 9
The recognition result stored in is displayed on the right side of the display 11, and the layout screen created by the layout creating section 14 is displayed on the left side (33). That is, on the screen of the display 11, the recognition result for one form or one line is displayed on the right side, and the layout for one form is displayed on the left side. Here, if you can not display all recognition results,
It goes without saying that a part of the recognition result is displayed and scrolled according to the instruction. Furthermore, since the display area of the recognition result is reflected on the layout screen, it is possible to perform modification to improve the operability by changing the color tone and the brightness of the display area. It goes without saying that the mechanism for changing the color tone and the brightness of the display area can be extremely easily configured from the conventionally known heart and software. For example, see Hitachi Personal Computer B16's BASIC.V2 explanations P 115-P 119.

When the operator sees this display result and finds an error or a reject character in the recognition result, the correction signal is given to the display correction controller 15 by pressing the keyboard 12. The display correction controller 15 rewrites the display result on the display 11 and also correctly rewrites the contents of the result buffer 9 (34). In this way, the operator can easily refer to the correct character on the form from the position information by looking at the layout screen, so that the operator can easily correct it using the key of the keyboard 12 corresponding to that character. It becomes possible.

FIG. 4 is a block diagram showing another embodiment according to the present invention.

FIG. 4 shows an example in which a storage device 16 is further connected to the display correction controller 15 as compared with the embodiment shown in FIG. In this example, each time the form is read, the form is read by the operator one after another without correction, and the recognition result and the character code of the candidate character required for correction and the ID number of the format information are stored in the storage device 16. Next, after the processing of a plurality of sheets is completed, the contents of the storage device 16 are read out and displayed on the display 11 and corrected by the operator. As described above, in this embodiment, the batch correction is possible, and the use efficiency of the system is remarkably improved.

As described above, according to the present embodiment, since the input form image is displayed by using the format information which is the code information, a large amount of character image data is transferred via the low speed serial interface between the OCR and the personal computer. It is not necessary, and the real-time character recognition result can be corrected. Further, since the layout screen is displayed as the result correction auxiliary screen, unlike the case of using the character image itself, the screen for one form can be easily displayed.

Further, this effect is remarkable in a large-sized form in which a large number of items are filled in by handwritten Chinese characters.

〔The invention's effect〕

According to the present invention, the recognized character pattern and the layout information of the character pattern can be simultaneously displayed in different areas on the screen, so that any part of a document (form) in which multiple items are described as a character pattern can be displayed. It is possible to easily determine whether or not the correction should be performed, and as a result, it is possible to easily perform the correction of the character pattern, and it is possible to significantly reduce the burden of the correction work on the operator.

[Brief description of drawings]

1 (a) and 1 (b) are views showing the principle of the present invention,
FIG. 2 is a block diagram showing an embodiment according to the present invention, FIG. 3 is a chart showing the flow of controller operation, and FIG.
The figure is a block diagram showing another embodiment according to the present invention. 1 ... Form, 2 ... Display 21 ... Form layout display area 22 ... Recognition result display area, 23 ... Recognition result display area 24 ... Candidate character display area, 3 ... Input form 4 ... Character observation Part 41 OCR main unit 5 recognition unit 6 result buffer 7 format storage unit 8 recognition controller 9 result buffer 9 1 personal computer 10 screen creation unit 11 display 12 …… Keyboard, 13 …… Format storage 14 …… Layout creation unit 15 …… Display modification controller 16 …… Storage device 101 …… Output terminal to external storage device

 ─────────────────────────────────────────────────── ─── Continuation of front page (56) References JP-A-58-189782 (JP, A) JP-A-55-157072 (JP, A)

Claims (2)

[Claims] 1. A character reading means for capturing at least a character pattern from a document having a character pattern and layout information surrounding the character pattern to obtain an image signal of the character pattern, and a storage means for storing the layout information on the document. Recognizing means for slicing out an image signal of a character pattern imaged by the character reading means based on format information obtained from layout information stored in the storing means, recognizing and storing the character pattern, Display means for displaying the layout information and the character pattern recognized and stored by the recognizing means in another area on the screen so that the layout information stored in the means can be referred to; and a display means for displaying on the screen of the display means. Character pattern recognition means for correcting at least a part of the generated character pattern. Correction device. 2. The display means is configured to highlight a range in the display area of the layout information corresponding to a range of the character pattern in the display area of the character pattern. The character pattern recognition and correction device according to item 1.